Two months ago, Asus announced the ROG Equalizer, a power cable for certain Nvidia RTX graphics cards designed to banish the fear of melting connectors once and for all. Well, perhaps not completely so, because it seems we now have the first case of a melted ROG Equalizer cable.

As posted on ChipHell and reported by Uniko's Hardware on X (via Videocardz), there's admittedly very little information to go on, other than a crystal-clear picture of a seriously fried connector at the end of an Asus ROG Equalizer cable. Three of the six power pins look properly scorched, with one showing a considerable degree of melting in the plastic housing.

However, there's no indication as to what PSU and graphics card this cable was plugged into, nor the circumstances as to how the melty-melty all came about, so the one thing we can't do right now is point our fingers at Asus and go 'Ha! Your cable is rubbish!' Well, we can, but we just can't use this solitary picture as evidence for such a judgment.

If you're unaware of what all the hoohar is about Nvidia's 12VHPWR (and the subsequent 12V-2x6 design update) connector, let me give you a quick explainer. In order to have its graphics card use gopping great globs of power but also not require lots of big connectors on the card's PCB, Nvidia designed a compact system to be used instead of the traditional 8-pin PCIe system.

A precursor to 12VHPWR first appeared on certain RTX 30-series cards, such as the RTX 3090 Ti and 3060 Ti Founders Editions. This design just comprised twelve power pins (six +12V, six ground), but the iteration of the design that was approved by PCI-SIG and adopted into the PCIe specifications also housed four more pins for sensing.

That format first showed up on Nvidia's RTX 40-series, and almost immediately, reports of melting connectors appeared from owners of GeForce RTX 4090 cards. So much so that it resulted in a class action lawsuit against Nvidia being filed in 2022.

The problem is quite simple. A full load, an RTX 4090 can use up to 450 W of power, but distributed across six pins, each one only has to deal with 6.25 A of current. However, should one pin fail in some way, or the graphics card draw more current in one or two pins than the others, the resulting load imbalance can push the current over the 12VHPWR's design rating of 8.33 A per pin.

Fast forward to today, and we now have a power supply market replete with all kinds of 'solutions' to the connector melting problem. Even the minor redesign, called 12V-2x6, did little to prevent the problem from occurring. Der8auer's detailed analysis of the Asus ROG Equalizer cable shows that while it's a decent enough product, it doesn't really offer much in the way of protection against the problem.

With RTX 5090 graphics cards now costing more than $4,000 in the US, you'd think there would be a far more concerted effort to banish meltygate once and for all. Part of the reason as to why we haven't seen a permanent solution yet is that the fact not every 4090 or 5090 owner has suffered from a melting connector.

The exact number of casualties is probably impossible to determine, but since graphics card vendors, as well as Nvidia, put money aside each year to cover warranty claims, the only way that they would really sit up and pay attention is if the total cost of all the replacements for damaged cards consistently exceeded the pot of money for claims.

We probably can't rely on hardware vendors coming up with a permanent solution, either. It's unlikely that Nvidia will abandon 12V-2x6 in favour of a completely new design, and having active load balancing on graphics cards will eat into a vendor's profit margins (though they're hardly small sums of money when it comes to the RTX 5090).

This particular instance of a melted ROG Equalizer cable could be entirely user error or even a clickbait thing, just to get a bit of attention, but since Asus' design has no means to actively prevent one pin from supplying too much current, this will almost certainly not be the only case we'll see, genuine or otherwise.

If you do have a very high-power RTX card with a 12VHPWR/12V-2x6 connector, here's my advice for keeping it as melt-free as possible:

1. Try to use a PSU that has a native 12VHPWR socket and cable: this won't guarantee perfect safety, but it removes the multiple points of failure you can get with adapter cables.
2. Don't keep removing the cable to check that it's okay: Power cables aren't designed to be repeatedly unplugged and jammed back in, and doing so can weaken the connections, leading to smaller contact patches and thus higher temperatures.
3. Explore undervolting your graphics card: You can just drop the maximum power limit for your RTX GPU in the Nvidia app, but you're better off following a detailed guide on how to lower your GPU's voltage across its range of clock speeds. If it all works as intended, you should get a graphics card that not only uses less power but is also cooler and, in some cases, able to clock higher.

Until we have graphics cards and/or power supply units that actively monitor the current draw on each pin, and dynamically prevent it from exceeding a set limit as standard, the melting connectors are still going to be a thing that will hit the headlines from time to time. Just like this one.

Computex 2026 has been a peculiar affair, primarily because the PC market and general global economy is just so shoddy right now. In addition to much of the tech on show seeming more out of reach than usual, though, this year's computing expo has shown me just how ephemeral the PC gaming market has become. Like a cloud, it certainly exists and you can see it from afar, but get up close and its boundaries start to become more obscure.

Obviously a lot of what we've showcased here at PC Gamer is stuff that's more exciting for PC gamers, but there's been plenty else at Computex that seems to not fit neatly into that 'just for PC gamers' category. And I know one could say that about pretty much anything in this space because most PC gaming gear can be used for other things, but it seems like even manufacturer intentions and target markets are now becoming a little blurred.

For one, we've had definitively non-PC gaming products that might actually be decent for PC gaming. Take Asus' new ProArt monitors, for instance. These are nominally for design professionals, but I don't think you'd find any gamer complaining about a 240 Hz QD-OLED.

Or there's the new Dell XPS 13, which seems to be the company's answer to the Apple MacBook Neo. It's aimed at students and professionals that use a laptop for light work, much like the Neo. However, its gaming chops are expected to be decent, too, thanks to its Wildcat Lake and before long its Panther Lake processor.

Even out of the products we've seen that are more specifically gaming-oriented, though, there have been some that fit into the 'not just for gaming' mould.

PC case line-ups often have unassuming and professional-looking options, so I suppose it's no surprise that the same is true at this year's Computex. In particular, the Tryx Vas caught my eye, with its metallic briefcase design and giant side fan. I could just as easily imagine this in an executive's office as a gamer's den.

Then there's Keychron's ultra-thin magnetic switch keyboard. I mean, tell me you like the Apple Magic keyboard without telling me you like the Apple Magic keyboard. This thing is practically crying out to be slapped on a Starbucks coffee table by a freelance digital nomad. But it's also magnetic, meaning it must surely benefit from all the analogue gubbins that gamers love, like rapid trigger.

And then there's Corsair's new TC80 gaming chair. The company explicitly distances itself from gaming here: "Following our popular TC500 Luxe, we have continued the move away from traditional gaming chair style. The TC80 has more subtle branding, a muted color pallet, and a more professional vibe overall."

It's still a gaming chair, but moving away from "traditional gaming chair style" means it's also kind of not. Or, rather, it's something that won't look or feel too out of place in an office but is still aimed in large part at PC gamers.

It's a trend that I've noticed for at least the last couple of years. I remember, for instance, when SteelSeries showed me its Apex Pro Gen 3 keyboard. During my trip to SteelSeries HQ in Copenhagen, the company was keen to emphasise that gamers are no longer just using their peripherals for gaming—they're using them for working, relaxing, travelling, and everything else in-between.

So SteelSeries was going to start catering for that a little, toning down the gamey-ness and making things a little more well-rounded. That seems to have been a general direction for many manufacturers over the last couple of years, though I'd probably call it companies cottoning on to the reality that was always the case: people use their products for multiple things, not just gaming.

The as-yet unmentioned elephant in the room, though, is of course Nvidia's new RTX Spark processor. The N1X-based chip was long-anticipated but only finally announced at Computex a couple of days ago.

The main take-home is that it's a chip with integrated graphics that promises some serious gaming performance despite not having a separate, discrete GPU. If Nvidia's RTX 5070-level claims hold up, it should even blow away the chips that sit inside handheld gaming PCs when it comes to gaming performance—but it's still all integrated. It'll be very game-dependent, given it's Arm-based, but in theory it could be fantastic for lightweight, efficient gaming.

And that last part is key: it should allow for lightweight and efficient devices—devices such as the RTX Spark-powered MSI Prestige N16 Flip AI+ that was announced alongside it at Computex.

RTX Spark is, of course, designed to be put inside machines you'll run AI on—this is 2026 Nvidia we're talking about, after all—but it's for gaming, too. In other words, RTX Spark machines are unlikely to be marketed as just gaming laptops or just productivity and AI machines.

This melding of gaming and non-gaming markets isn't completely new, but witnessing this year's Computex really brought it home. The lean into do-it-all products seems one that's here to stay. And perhaps that's a good thing, with the market being as it is—when things are so expensive and the cost of living is so high, I'd rather spend my money on one device or peripheral that does everything well.

The old marketing saying of 'sex sells' frequently gets converted to 'excess sells' when it comes to computer tech, and for the 20th anniversary of the ROG brand, Asus has taken this to another level with a new 3000 W power supply unit for gaming PCs.

It's not the power rating that's so eyebrow-raising about the Asus ROG Thor 3000W Titanium III Edition 20, as 3 kW PSUs aren't anything new in the world of workstations. For example, Corsair has the WS3000 and Asus itself has the Pro WS 3000W Platinum. However, for gaming PCs, it's completely overkill and totally unnecessary, even if you're overclocking the twangers off every single component in your rig.

And yet, that's the exact market that Asus is selling its latest mega supply to. Making matters even worse is the fact that the PSU sports four 12V-2x6 power sockets, despite the fact that you can only game on a single RTX 5090. Sure, you could slap another three into your PC for giggles, but it won't help to make anything run faster.

For AI workloads, it's a different story, but that's not mentioned in Asus' marketing, nor is it what ROG is all about. The Thor 3000 W Titanium III Edition 20 does have gaming-related features, such as an OLED display that can be magnetically attached to the PSU or your PC's case via an extension cable, plus a voltage monitoring system to prevent your RTX 5090 from self-destructing.

But no gaming setup is going to get anywhere near the unit's 3000 W power limit (1600 W when using a 115 V supply), and even though the Thor's peak efficiency will be around the 1600 W mark, that's still far more than you're likely to need.

Well, unless you just happen to have an overclocked 800 W Asus RTX 5090, plus a heavily PBO'd Ryzen 9 9950X3D2 Dual Edition, a mountain of DRAM sticks, an even bigger mountain of Gen5 SSDs, plus a small warehouse-worth of fans and RGB lighting.

Of course, I am applying logic to a product that isn't logical. It's not really a gaming PSU at all: It's an over-the-top statement from Asus, simply because it can, and partly because the whole ROG brand has become synonymous with this line of thinking.

But even if you can afford it, I would recommend that you look elsewhere for an all-singing, all-dancing power supply, because when you look closely at Asus' images for the new PSU, you can see that it's just a spruced-up standard design. The main power switch and some of the connectors also look somewhat wonky, indicating that this isn't a special, hand-crafted unit. Ho-hum.

The OLED gaming monitor tech announcements are coming thick and fast at Computex this year. One intriguing new monitor has been outed by Asus featuring a new WOLED panel from LG that we've not seen before.

The big news is that the Asus ROG Swift PG32UCWM features RGB-stripe technology. In other words, the panel's subpixel structure is composed of conventional red, green and blue vertical subpixels in that order.

Until recently, all OLED gaming monitors had panels with various weird and wonderful subpixel structures. LG's panels had an extra white subpixel to boost peak brightness. Samsung used a triangular layout, albeit with the standard RGB colours.

These non-standard subpixel layouts cause various issues. So, the ideal solution is an absolutely standard RGB-stripe structure. And that's what the Asus ROG Swift PG32UCWM's new LG WOLED panel offers.

Note that the "W" in LG's WOLED branding doesn't refer to a white subpixel, but the white colour of the base emissive OLED material, which is then passed through red, green and blue filters for each subpixel, respectively.

Anywho, the ROG Swift PG32UCWM is, as I said, a 32-inch 4K model. It runs at 240 Hz at 4K and offers an alternative 1080p mode at 480 Hz. It also sports VESA DisplayHDR True Black 400 certification. And that's where things get a bit tricky.

You can already buy 32-inch 4K Samsung QD-OLED based monitors, including the MSI MPG 322UR X24 I reviewed recently, that offer brighter True Black 500 certification. And Samsung has announced yet another 32-inch 4K panel with True Black 600 certification, 360 Hz refresh and RGB-stripe subpixels for later this year.

That latter panel in particular makes the new LG WOLED in the Asus ROG Swift PG32UCWM look a bit behind the curve, even as it launches. It'll do 350 nits full screen and 1,500 peak HDR.

The LG WOLED in the Asus, meanwhile, tops out at 1,000 nits peak HDR. Asus hasn't quoted as fullscreen figure. But it will likely be in the 250 to 275 nits range. These brightness figures are despite the panel using LG's latest Tandem OLED tech. And that's likely because of the removal of the white subpixel, which LG has thus far used to increase the brightness rating of its OLEDs.

Still, the good news here is that OLED technology is really kicking on this year. We're getting brighter panels seemingly by the week and RGB-stripe is being ever more widely adopted.

I wouldn't be surprised if most OLED gaming monitors are RGB-stripe by this time next year. Add in increased brightness and we're surely approaching a situation where the art of OLED gaming monitors is pretty much perfected.

Well, maybe with one final caveat around burn-in remaining. But the evidence continues to mount around existing panels that burn-in isn't a massive issue. And these newer panels are meant to be yet more robust.

We've seen all manner of keycaps over the years, from Kerbals to desserts and just about everything in between. But I don't think I've seen any quite so peculiar as those just announced from Asus ROG.

They come in a 'Mystery Box Edition 20', with the latter part of this referring to the company's 20th anniversary line-up. We've already seen an absolute chonker of a GPU under this moniker, so I guess it makes sense to have something small in the range too like keycaps.

It looks like they are literally mystery boxes, in which you don't know which keycaps you're getting. In one box you get 6x mystery keycaps. And what are the keycaps, exactly? Well, it's all ROG peripherals, of course.

Each keycap sits on top of a bed of stalagmite crystals. To start with, there's a Gladius mouse, a Raikiri controller, an Asus Delta headset, alongside a couple of different versions of the ROG logo.

The last couple are the most interest of the bunch, though. There's the Chariot chair which, erm, spins around. Not on its own, you have to spin it yourself, but still. Peculiar. Then there's the Claymore keyboard which is split into two. The left side has the bulk of the keyboard, while the right has the numpad, and both can be pressed independently. Y'know, just 'cause.

The only thing that I think would make it better would be if there were a little keyboard keycap on the keyboard keycap. But alas, we're just a single order of magnitude into bonkers territory here.

It's a bit silly, but it's also just a bit of fun, so I'm not too against it. The company has announced plenty of more serious products at this year's Computex, including an esports OLED monitor I quite like the look of, so I'll forgive Asus for the frivolity here. Though much will depend on just how rare the fancier keycaps are, because I wouldn't much fancy having to rack up dozens of spares just to collect them all.

With Asus' ROG brand turning 20 years old this year, it has announced a whole pile of fancy black and gold gaming gear at Computex. Its new 82 g mouse is just as fancy as you'd expect and comes with some truly outrageous specs.

Called the ROG Harpe II Extreme Edition 20, Asus's new gaming mouse has "Authentic 24K gold-plated detailing", as well as a black translucent shell. The first time I saw it, I thought it looked a bit like it had cardboard internals (because gold under black certainly muted the colour palette a tad), but I don't think it looks bad by any stretch.

With it, you get a max 8,000 Hz polling rate (both wired and wireless), along with Asus's own ROG 100M optical micro switches, and a 65K DPI sensor. Those are truly absurdly high specs. The Razer Viper V4 Pro, our pick for the best gaming mouse, also goes up to 8,000 Hz polling but only 50,000 DPI. However, the max polling and max DPI of both of these gaming mice are arguably too much for 99% of players, and 8k polling often isn't worth the battery drain for a wireless mouse.

82 g is a good bit heftier than the V4 Pro, and I suspect that could have something to do with the gold internals. It is based on the Harpe II Ace (which is only 48 grams), and Asus says its shape was developed with the help of esports pros.

Asus also notes it comes in a "display-ready presentation box", and that box seems to come with stickers, a cable, the connectors, and a keyring of some kind. Asus has also announced the ROG Azoth Extreme Edition 20, which is a black and gold-themed variant of the ROG Azoth gaming keyboard.

Asus has confirmed to me that the Harpe II Extreme Edition will be $260 and the Azoth Extreme Edition will be $600. That's a lot more than the Viper 4 Pro, and Razer's offering isn't exactly cheap.

As with all of Asus' ROG 20th anniversary items, this gaming mouse seems like a commemorative item, more so than something you'd just buy to simply use. And I'd imagine you'd need to have quite a bit of cash burning a hole in your pocket to consider it. Here's looking at you, Xbox Ally X20 bundle.

As hardware goes, graphics cards can be on the heftier side. Asus ROG's Astral cards were already a case in point, but the company has just unveiled an anniversary edition well-deserving of the two C's in 'thicc'.

During this year's Computex, Asus announced the ROG Astral GeForce RTX 5090 Edition 20. The GPU isn't only CHONKY, but makes the most of its expanded surface area with a curved AMOLED display and "a quad-fan design that delivers up to 20% improved airflow." The main drawback is that this big boy graphics card will likely gobble up not only the contents of your wallet, but all of your PCIe slots too.

The Edition 20 series celebrates the ROG (Republic of Gamers) sub-brand's 20th anniversary. The full ROG Edition 20 lineup encompasses a number of other PC components, including a mother board and even an NUC, marking all of the hardware with a black-gold colourway. Asus Chairman Jonney Shih said during his opening speech at Computex 2026, "Twenty years on, we didn't just change the game, we transformed the world around it."

No firm release window for the ROG Astral GeForce RTX 5090 Edition 20 has yet been announced. So far, Asus has only said, "Many limited-edition items and the exclusive Armoury Crate Edition 20 software theme will be released as limited-time offerings throughout 2026."

Pricing details are also thin on the ground. However, judging by the ROG Astral GeForce RTX 5090 most recent flirtation with all that glitters—specifically, the $7,000 gold-plated Asus ROG Astral GeForce RTX 5090 Dhahab OC Edition—it will likely cost a pretty penny. The slightly less powerful Asus ROG Astral RTX 5080 Dhabab Core edition still went for $2,589. With those kinds of numbers, it hardly seems worth saving up my pennies.

This card, along with the other ROG Edition 20 components, is limited edition, just as last year's ROG Matrix RTX 5090 was. There were only 1,000 models of that one, and of course limited supplies tends to mean higher prices.

Still though, the ROG Astral GeForce RTX 5090 mounted on the corresponding Edition 20 the open-frame ROG GR20 PC case was a thing of beauty to behold on the Computex showfloor. I'd be inclined to consider an open concept set up like this show piece were it not for the fact my home office is an absolute dust magnet—and I'm not made of PCIe slots, either.

The Asus ROG Xbox Ally X is undeniably a great device, but it's an expensive one. So when I saw Asus had unveiled an upgraded anniversary edition, bundled with a pair of AR glasses, I felt a little pain for my bank account.

The Asus ROG Xbox Ally X20 bundle includes an Xbox ROG Ally X with a snazzy black and gold theming (including a 'custom gold internal structure'), TMR thumbsticks, and even the addition of a 7.4-inch OLED screen.

For context, the standard ROG Xbox Ally X comes with a 7-inch IPS LCD screen, so this is already a major upgrade. Both screens have a 120 Hz refresh rate. The TMR thumbsticks are a welcome addition, meaning you're less likely to have to take the thing apart due to stick drift. TMR sticks should also make for a slightly better battery life, too.

Finally, it comes with a new D-Pad that can be swapped between four-way and eight-way input.

In the new models, you can expect almost the same processor (the Ryzen AI Z2 Extreme in the X20 has an NPU, but other than that, it's exactly the same as the Ryzen Z2 Extreme), which means pretty much identical game performance, as well as 24 GB of LPDDR5X memory, a 1 TB SSD, and an 80 Wh battery. The X20 is around 40 g heavier, but everything else remains the same.

Unfortunately, we don't know the answer to the biggest questions here. We don't have a confirmed price point or a release date. There are some random numbers on the side of the page when you get to the 20th anniversary section, so they could correlate, or they could… not.

As for price, I'm expecting a very heavy hit here. Not only are you getting some meaningful upgrades on the original $1,000 Xbox Ally X, but you're also adding in a pair of ROG XReal R1 AR glasses, which fetch around $1,000 by themselves. This is all before mentioning that the ongoing memory crisis has been a major pain when it comes to reasonably priced hardware, and 24 GB of memory is quite a lot.

The sell for this X20 bundle is the pairing of the glasses and the handheld. You can plug the 240 Hz glasses into the top and get what appears to be a 171-inch screen right in front of you. It is tethered to the device, though, so you likely won't be whipping out that bad boy on the train. Given what I expect its cost to be, I wouldn't want to bring it out of the house anyway.

The march of OLED technology continues. According to market research outfit Trendforce, OLED monitor shipments are up by 78% year-on-year in the first quarter of 2026. Trendforce reckons increased supply of Samsung's QD-OLED panels is a key factor, but it's actually Asus that is selling the most OLED monitors.

Q1 2026's figures are actually down 11% on Q4 2025. But given tech products have distinct annual cycles driven by things like school schedules and holiday periods, the most representative comparisons are with the same quarter the previous year. In that context, OLED's continuing progress is clear enough.

Apart from the broader trend toward OLED, Trendforce reckons Samsung ramping up supply is a key factor. "The key growth driver was the increasingly abundant supply of QD-OLED panels, which enabled new market entrants to ramp up volumes and effectively fill gaps in the market," Trendforce says.

Perhaps somewhat surprisingly given its premium pricing, it's actually Asus that is selling the most OLED monitors, with 24% market share. "Its leadership was underpinned by a comprehensive OLED product lineup," Trendforce says, noting how broad Asus's OLED monitor range is.

Samsung (Samsung Electronics that makes monitors, as opposed to Samsung Displays that makes the actual QD-OLED panels) came second with a 16.4% share, while MSI clocked third with 12.2%.

This time around, Trendforce isn't quoting overall volumes of OLED monitor shipments. But to put all this into context, it previously said that the overall OLED monitor market amounted to 2.735 million units in 2025, a 92% increase on 2024.

While it's a somewhat subjective call, I'd say Samsung driving volumes of OLED monitors is a good thing. Samsung's QD-OLED panels tend to be more vibrant than LG's competing WOLED panels thanks to utilising a pure RGB subpixel structure where LG relies on an additional white subpixel to boost measured brightness in a way that doesn't always translate well into the real-world experience.

Put another way, Samsung's latest QD-OLED panel tech, as seen in the MSI MPG 341CQR X36 in 34-inch ultrawide form and the MSI MPG 322UR X24 in 32-inch 4K configuration, is pretty darn stellar.

When Asus took over Intel's baton for the Next Unit of Computing (aka the NUC or mini PC), some of us hoped that we'd be treated to a fresh new direction to the original concept. Essentially, that's what we got, but it also came with something we didn't want, and the latest Asus ROG NUC perhaps embodies everything that has gone wrong with the whole platform.

Ostensibly, there's not a great deal of difference between the freshly announced ROG NUC 16 and last year's ROG NUC. They both sport Intel Arrow Lake HX processors (albeit the Plus models in the 16) and a GeForce RTX 50-series mobile GPU; you can fit up to 128 GB of memory (up from 96 GB) and two M.2 SSDs in both of them, as well.

The looks have been spruced up, with the ROG NUC 16 being a tad flashier in appearance than the older design. It's not a radical change, and if I'm permitted to be honest, the white model (sorry, Moonlight White Edition) gives me 'Temu PlayStation' vibes. But mini PC aesthetics are very much a subjective thing, and if it's thumbs up for you, then good on ya.

There's no indication of the official MSRPs for each variant of the ROG NUC 16 just yet, but Videocardz reports that pre-orders in China are roughly in the $4,400 ballpark. Right now, if you're so inclined, you can bag yourself the RTX 5080 version of the previous NUC, with 32 GB of DRAM and 2 TB SSD, for a mere $3,868 at Amazon.

So it looks like if you want a top-spec ROG NUC 16, you'll need to find a spare four thousand bucks stuffed down the back of your couch. Before you splash the cash, though, you should know that this amount of money will get you a Lenovo Legion Pro 7i with the same hardware and still leave $1,000 spare for an extra display and peripherals.

NUCs-for-gamers, and mini PCs in general, are always going to be a niche product and thus a low-volume seller, so an inflated price tag is to be expected. But this? Four kilo dollars is beyond any measure of reason or justification, despite the grimness of the RAMpocalypse. Sure, the new NUC can be kitted out with up to 128 GB of DDR5-6400, which you might think would be great for AI modelling.

However, Asus is marketing this device at gamers—it claims all that memory will ensure "smoother high-FPS gameplay and faster responsiveness"—and the NPU in the Arrow Lake processor isn't good enough to make proper use of masses of DRAM; for that, you need something like the Framework Desktop. And if you use the GPU instead, you'll be limited to a maximum of 16 GB of VRAM (RTX 5080 variant).

Other than taking up considerably less desk space than an equivalent laptop, the Asus ROG NUC 16 makes zero sense as a gaming purchase, or any kind of purchase for that matter. Mini PCs are supposed to be ultra-compact, with their compromised performance offset by an enticing price tag.

Asus does make such NUCs, though they're aimed at the business and education markets. What we've got here is less something for PC gamers, and more for cash cows that can be won over with ROG branding and top-tier hardware specs.

It's the 20th anniversary of the Republic of Gamers (ROG) and although we've already been treated to a special edition copper-themed Crosshair motherboard, it looks like Asus has something extra lined up for PC Gamers: its own line of DRAM kits.

Well, strictly speaking, it's just the one kit so far, and even then, it might just be a one-off, as part of the ROG celebration. Videocardz reports that Asus dropped the deets on the DRAM at its ROG Day 2026 event in China, but for everyone else, there's a preview look at the DDR5 set via 林大餅Bing's YouTube account.

Apparently developed via a collaboration with Biwin, the 48 GB dual-channel kit is rated to 6,000 MT/s with CL26-36-36-76 timings. That CAS latency is very tight, and I dare say that a good number of AM5 motherboards won't like it very much. Naturally, you can expect it to be fine with Asus' high-end boards, AMD and Intel flavoured, as it comes with XMP profiles for both platforms.

In terms of looks, well, it's ROG-a-go-go, as you'd expect, but I have to say that I quite like the design: angular, but not garish, with a rather classy colour scheme.

As for the price, Videocardz says that it's listed as 5999 yuan, which directly converts to a fraction over $888. How much of that is down to the RAMpocalypse is anyone's guess, but a 48 GB set of Biwin DDR5-6000 CL28 is $540 at Amazon. Specialised DRAM kits are always pricier than the regular stuff, so once you account for that (plus a bit extra for the ROG tax), I reckon we'll be looking at around $850 retail.

The interesting thing about all of this is whether or not Asus has been planning on entering the desktop memory market en masse. If this DDR5 kit is the only thing it's going to release in this sector, then the specs and pricing make sense, but if it's part of a long-term strategy to announce being a competitor to Corsair, G.Skill, and many others, it couldn't have come at a worse time.

ROG enthusiasts won't really care, though, and will probably try to snap up a set as soon as they can. After all, the only part of a complete gaming PC setup that Asus currently don't offer is a CPU and a memory kit. Oh, and internal SSDs, though you can get external units.

But if it's happy to start producing its own branded DRAM, I wouldn't be surprised if a couple of M.2 drives suddenly appeared, perhaps even followed by an Arm-powered processor by the end of the decade. Anyone fancy a 100% ROG rig?

Even though Asus is doing its very best to let us all know that this year marks the 20th anniversary of its Republic of Gamers (ROG) branding, you may well not have picked up on the fact that it's releasing all kinds of one-off goodies to celebrate two decades of high-end gaming gear. But whatever else Asus has got planned, it's not going to top this pure blast of nostalgia for me.

Fellow PC enthusiasts, I give you the Asus ROG Crosshair 2006. Yes, I know it's just a motherboard, and if you're a recent starter in one of the best hobbies around, you might not think much about it. But if you were a hardware enthusiast around the start of the millennium, then you'll know exactly why this is ticking every box for me.

Back then, every PC component aimed at the hardcore crowd was plastered with copper heatsinks, often in some ludicrously over-the-top form that was unnecessary, since power levels back then weren't anything like as high as they are now.

To that end, Asus took its very first ROG motherboard, its latest ROG X870E Dark Hero model, and blended the two in what can only be described as a total smorgasbord of copper. Alas, it's not real copper. For example, the VRM heatsinks are "lightweight aluminum" that have been given "an eye-catching copper hue."

That's pretty disappointing, truth be told, but not entirely unexpected. After all, copper is horribly expensive these days, and that amount of the metal would make the motherboard very heavy and absurdly pricey.

And while copper has one of the highest thermal conductivities of any pure metal (second only to silver) and roughly double that of aluminum, the latter is more than good enough for use in today's cooling contraptions.

I don't really care, though. To me, the Crosshair 2006 is all about remembering the grand old times of PC modding, tweaking, and messing about. I'm not suggesting that it was better or easier to do back then; it's just that there were bigger gains to be found from overclocking stuff compared to what you can typically get with today's hardware.

This motherboard isn't a return to those days, just a visual reminder of my glory days of PC hardware schenanigans, and there's nothing wrong with that.

A new report suggests motherboard manufacturers are in a bad state right now, with the memory crisis and the Iran war affecting all kinds of PC hardware. As PC builders are disincentivised from, well, building PCs, the motherboards needed for them also get left behind.

The report, from Digitimes, notes a "collapse" for many motherboard manufacturers, claiming it is worse than previous financial crises and the first year of the Covid-19 pandemic. It notes that MSI, Gigabyte and ASRock are all expected to see declines in 2026, in regard to motherboard sales. As of the time of writing, motherboard sales figures for 2026 are not yet public.

The first element of this collapse is the memory crisis. As AI is intent on sucking all that memory for it to run, those looking to buy new DRAM kits are now paying a steep premium. Some simply won't and will hang on for it to go down. This memory crisis also hits other elements of a PC, like graphics cards, making them harder to buy at MSRP.

Jese Martinez from custom PC company PowerGPU recently affirmed the same to me, arguing, "it's memory, it's storage, it's multiple things that are happening. And it's not just a few 100 bucks. Stuff is going up twice the price or three times the price."

Naturally, if the price of components is stopping people from being able to upgrade, there's no point in buying a motherboard for the day it eventually goes back down. If the crisis lasts beyond 2027, one might assume they'd prefer to wait for the latest thing anyway.

However, that's not the only factor in this crisis. Digitimes notes that interest in Nvidia's RTX 50 series line of graphics cards is lower, too, due to a lack of updates and upgrades since the start of the year. We haven't seen the rumoured RTX 50 Super series yet, and it's unclear if we will, either. The report argues, "The mid-to-high-end gaming PC market lacks technical specifications that stimulate upgrades."

Digitimes claims that Asus, the industry leader in motherboards, sold 15 million units in 2025, but shipped just 5 million in the first half of 2026, indicating a steep drop for this year. It goes on to suggest that Asus is aiming for 10 million in total for 2026, which is apparently the lowest number of shipments in over a decade.

This doesn't necessarily mean the companies are in trouble, though. Asus' March earnings call argues its high-end brand value and shift to AI would help it weather the memory crisis. Gigabyte and ASRock, similarly, are looking to make money from AI. Though AI is arguably the biggest factor in the downturn of the motherboard market, it is also a money-making machine, responsible for Nvidia being worth $5 trillion. The main group that will be affected by this are the buyers, but AI doesn't seem to really care about them.

With the vast majority of handheld gaming PCs using an AMD chip to handle graphics duties, users have been mostly limited to using FSR upscaling to give frame rates a healthy bump. However, with its latest update, the Asus ROG Xbox Ally X now offers Microsoft's Auto SR as a different option, and this upscaler is powered by the processor's NPU.

You know what an NPU or neural processing unit is, right? It's that thing inside some processors that seldom gets used for anything. Don't get me wrong, it's a neat bit of hardware that's great for handling AI models without using masses of power, but hardly any applications require one to be present.

But with the Auto SR Preview feature in the new Asus ROG Xbox Ally X update, that's no longer the case, because Microsoft's upscaler doesn't run on the GPU inside the AMD Ryzen Z2 Extreme that powers the handheld PC. Instead, it's all done on the chip's NPU.

Microsoft explains why this is a good thing in a short blog about AutoSR: "When super resolution [upscaling] runs on the GPU, it counts towards frame time. To avoid impacting FPS, models are limited to a minuscule 1–2ms, constraining their size and quality. Game-integrated super resolution fits in this window and still delivers quality by relying on the game to provide more detailed texture data.

"Auto SR sidesteps this limit by running larger models on the NPU in parallel with the GPU. This gives Auto SR an entire extra frame of time to run the model—critical for devices like the ROG Xbox Ally X, that couldn’t otherwise run these models without significantly impacting FPS."

Now, it's important to note that Microsoft isn't saying that Xbox Ally X users should abandon using AMD's shader-based FSR in favour of Auto SR. Quite the opposite, in fact: "Game-integrated super resolution remains the preferred choice. Auto SR steps in where game-integrated super resolution isn’t available or when hardware constraints prevent it from simultaneously delivering quality and FPS."

In other words, a game that has been developed to use FSR natively is likely to produce better results than what Microsoft's automatic upscaler can achieve, but if that's not an option in the game, then Auto SR comes to the rescue. Well, sort of.

The system only works when the Xbox Ally X is docked and connected to an external display. The latter isn't a requirement; it's just that the only reason why you'd game on a docked handheld is to max out the power limit, so you can get the best frame rate on a big screen. But this does mean that you're still going to be using FSR, or AMD's driver-based RSR, when using the device in handheld mode.

That suggests the NPU has to work pretty hard to do Microsoft's upscaler, and even though it's designed to operate in low power conditions, the combination of the CPU, GPU, and NPU's power consumption probably isn't great for unplugged battery life.

And unfortunately, even with the update for Asus' handheld, getting Auto SR onto your Xbox Al…no, I'm not writing that silly name out again. Getting Auto SR onto your XAX is somewhat of a faff. You need to enrol in Xbox Insider to begin with, go through all the required updates in Windows, open the Xbox Game Bar to check that Auto SR is available, and install the latest Auto SR package from the Microsoft Store.

Let's hope that's all a lot simpler and easier to do once it's out of preview. Anyway, my take on all of this is that Auto SR will be a nice extra to have on laptops, once it's rolled out for more devices. And when desktop PCs with Intel Core Ultra chips, for example, get added to that pile, it might be a great way to finally make use of that NPU, while giving your graphics card a bit more breathing space to render your favourite games.

The ASUS ROG Xbox Ally released in late 2025 and was marketed as an Xbox device (remember This Is An Xbox?), but it'll play mostly everything in your Steam library—providing the hardware specs allow it—and it even plays nice with the Epic Games Store. It was the lesser spec'ed of two Xbox Allys: the ASUS ROG Xbox Ally X is the more premium model and PC Gamer's Dave James reviewed it favourably, though it is extremely expensive, especially in light of a February price hike.

Not wanting to fork out AU$1,799 for the ROG Xbox Ally X is totally understandable, but if you're after a decent handheld that'll play the vast majority of indies plus a lot of modern blockbusters, the ASUS ROG Xbox Ally will probably suit you. Its launch price was AU$999, but right now Amazon is selling it for AU$769, shipping included.

It's not the cheapest handheld gaming PC on the market: that would be the OG Steam Deck LCD, which Valve is selling for AU$649 at the moment until stock runs out, at which point it'll be gone for good.

So which should you get? I love the Steam Deck, and have written about how indispensable its trackpads have become for me. Performance between Steam Deck LCD and Xbox Ally are pretty similar, though the Xbox Ally has a 7 inch 1080p display with a 120Hz refresh rate, while the LCD has a 7 inch 800p display. The Xbox Ally has double the storage of the Steam Deck LCD at 512GB.

Leaving performance aside, the hugest point in the Xbox Ally's favour is that it runs on Windows 11. That's not good for a lot of reasons, but it's ideal for players who predominantly play online games that implement anticheat that's incompatible with Steam's Linux-based OS, such as Marathon or Destiny 2. The Xbox Ally is a clear winner in that department.

If you're in the market for a handheld gaming PC and don't mind spending more, check out our roundup of the best handheld gaming PCs.

Microsoft's fabled Project Helix chip, the AMD-engineered basis of the next-gen Xbox console, will reportedly be used by Asus, MSI and other brands to make third-party gaming machines. However, consumers won't be able to buy the Helix chip separately.

So says KeplerL2, one of the most reliable members of the tech-leak illuminati. Posting on NeoGAF forum (via WCCFTech), KeplerL2 said, "it won't be sold directly to consumers, but you will be able to buy an ASUS / MSI / etc Helix machine."

The idea that Microsoft will branch out from producing and selling only its own Xbox-branded hardware isn't new. Nor is the idea that an OEM like Asus would get involved. After all, we already have the economically monikered (not) Asus ROG Xbox Ally X handheld gaming device acting as something of a harbinger or early template for this approach.

However, the Xbox Ally X isn't a perfect model for the future living room console. It uses an essentially off-the-shelf AMD APU in the Ryzen Z2 Extreme, a chip that's not only available in non-Xbox handhelds, but also in everything from laptops to desktops, albeit under different branding as part of another tranche of the AMD Ryzen product range.

Project Helix, on the other hand, will be exclusive to Microsoft. Sure, it's engineered by AMD and most of not all the IP will be available in other chips. Currently, that's expected to translate into Zen 6 CPU cores and a GPU most closely related to AMD's next-gen RDNA 5 family of PC graphics cards.

But the precise mix of CPU and GPU core count—and perhaps some specific features in the ray tracing and / or machine learning spheres—may be exclusive to Helix. Moreover, console chips are traditionally built in big numbers and with few if any variations.

That tends to make them cheaper to make and deliver more bang-for-buck compared to traditional gaming PC hardware solutions, albeit at the cost of losing most if not all end-user configuration options. Console chips are what console chips do.

The catch, of course, is the possible confounding influence of console subsidies, where the expectation of games license revenues allows console makers to sell machines cheaper or even at a loss but still make a profit on the overall business.

Still, one of the interesting possibilities around the Helix chip has been the idea that if it were sold separately, it might form the basis of a powerful but relatively inexpensive PC gaming rig.

Exactly how Project Helix will be engineered isn't clear. But if it has on-package memory, the result would be a solution that just needed a motherboard to drop into and some storage and Bob's your next-gen-gaming parental sibling. Or, perhaps more likely, it will be a soldered-on chip with no socketing to speak of.

One might imagine, then, if Microsoft is banging its own-branded Xbox Helix boxes out for, say, $800 a pop with a fancy chassis and packaging, plus some controllers, some kind of bare bones setup, maybe using your existing case, SSD and PSU to cut the price even further then—hey presto—there's your $500 comprehensive gaming rig upgrade.

Well, however realistic that ever was, KeplerL2's post seems to consign it to history. It ain't happening because you won't be able to buy the Helix chip separately. That still leaves the likes of Asus or MSI to potentially undercut the official Microsoft console, however.

Given how pricey the Asus ROG Xbox Ally X is, Asus isn't a great candidate for that solution. MSI could be better, but even then, what with the memory and broader computer chip crisis, it would be a tall order for any OEM to hit a spectacular price point with a new device.

Arguably, Apple has just about achieved that with the MacBook Neo, though that obviously isn't a gaming device. But even Valve is struggling to cope with chip prices to the extent that the upcoming Steam Machine's launch has been disrupted.

All of which means that Project Helix doesn't immediately look like it will be the budget PC gaming saviour some observers once thought it might just deliver. It's still an intriguing chip and, personally, I can't wait to see what mix of AMD tech Microsoft has gone for. But I doubt it will end up being the basis of a new generation of cheap gaming PCs.

On paper, the AMD Radeon RX 9060 XT 8 GB makes a whole lot of sense. I was particularly impressed by the RX 9060 XT 16 GB when I tested it last year, and the 8 GB version is usually substantially cheaper. In fact, out of all of AMD and Nvidia's current-generation graphics cards, it's the most affordable of the lot. Dropping the VRAM to 8 GB can't make that much of a difference, can it?

Well, I hate to be the bearer of bad news, but yes it can. Not all the time, of course. Certain games at high settings, though, will hamper the AMD card's performance quite badly—and crucially, in a more significant way than its main competition, the Nvidia RTX 5060 Ti 8 GB.

Does that make the RX 9060 XT 8 GB a total write off? Nope. In fact, if the Nvidia card didn't exist, I'd be tempted to call the AMD GPU's performance pretty good for its paltry MSRP of $299. The RTX 5060 Ti 8 GB doesn't have it all its own way in my benchmarks, either. Sometimes the AMD contender pulls ahead, so it's not a super clear-cut thing.

But there are a couple of caveats to be aware of. One, you won't find one of these for less than $350 right now, and the RTX 5060 Ti 8 GB tends to bottom out around $390-$400. And two, when the RX 9060 XT 8 GB drops off the pace, it's sometimes by a significant amount. Again, these drops are in certain games at certain settings, so it's not a completely consistent thing. But it's enough to give me pause, and I think you should consider your options, too.

It's far from a bad graphics card, and one that'll serve you fine if money's a real concern and you need something that can cope with modern gaming. But the RTX 5060 Ti 8 GB handles VRAM-constrained situations significantly better in many of my tests, and that's the sort of headroom that I think's worth paying for. At least when the price difference between the two is this tight.

RX 9060 XT 8 GB - Features

The RX 9060 XT 8 GB uses the same Navi 44 chip as the 16 GB version, based on AMD's RDNA 4 architecture. With that comes a loadout of 2048 Shader Cores, 32 Compute Units, 32 dedicated Ray Accelerators, and 64 Matrix Cores—essentially the equivalent of Nvidia's mixed precision, AI inferencing Tensor Cores.

While both the RX 9060 XT 8 GB and the RTX 5060 Ti 8 GB have the same RAM loadout in terms of size, the Nvidia card uses speedy GDDR7 over a 128-bit bus for a total max memory bandwidth of 448 GB/s, while the AMD equivalent uses slower GDDR6 over the same sized bus. That means a max memory bandwidth of 322.23 GB/s, significantly less than the RTX 5060 Ti.

The RX 9060 XT has 32 KB of L0 cache per workgroup processor (essentially two Compute Units), 4 MB of L2 cache, and 32 MB of L3. This multi-layered system is again in contrast to Nvidia's cache handling, which equates to 128 KB per Shader Module of L1 cache, and 32 MB of L2.

In terms of clock speeds, the stock RX 9060 XT has a base clock of 1700 MHz, a "game clock" of 2530 MHz, and a max boost clock of 3130 MHz.

The game clock figure represents the average sustained speed during a typical gaming workload, although my particular review sample, the Asus Prime Radeon RX 9060 XT OC Edition, lists a higher sustained game clock figure of 2760 MHz—likely due to the fact it has a surprisingly large, tri-fan, 2.5 slot cooler.

This seems a little overkill for an RX 9060 XT, which has proven itself to be a cool-running graphics chip even with the small, twin-fan cooler equipped to the 16 GB version I tested last year. I found that card to regularly sustain somewhere near its top max boost speeds while keeping temperatures low, and this Asus 8 GB sample is much the same.

Asus doesn't list the exact TGP of this particular variant, but it does suggest a 550 W power supply, 100 W more than the official minimum spec. The default RX 9060 XT 8 GB has a TGP of 150 W, although I've recorded a peak power draw of this card at 175 W under extremely heavy load.

Again, the default RX 9060 XT 16 GB sample I reviewed previously also liked to jump over its 160 W TGP at points, managing an even larger peak of 182 W.

Around the back is a single HDMI 2.1b port, and two 2.1a DisplayPort connections. One last thing to note about this particular Asus card is the addition of a small switch tucked into the backplate, which allows you to switch between Performance and Quiet fan modes. I've left it in performance mode for my testing, as this card runs quiet even when pushed to its maximum.

RX 9060 XT 8 GB - Performance

And so, to the meat and potatoes: gaming performance. Initially, I was pleased to see that the cheaper RX 9060 XT 8 GB managed to get within three frames of the RTX 5060 Ti 8 GB in Black Myth Wukong at 1080p on average, which suggested to me that I was likely to have a fight on my hands between the two.

In Cyberpunk 2077, however, the wheels really come off for the AMD card.

1080p gaming benchmarks | RX 9060 XT 8 GB

Black Myth Wukong (1080p High)Cyberpunk 2077 (1080p RT Ultra)F1 24 (1080p Ultra high)Homeworld 3 (1080p Epic)Metro Exodus Enhanced Edition (1080p Ultra)The Talos Principle 2 (1080p Ultra)Total War: Warhammer 3 (1080p Ultra)

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

71

58

RTX 5060 Ti 8 GB Palit Dual

74

62

RX 9060 XT 16 GB XFX

73

63

RTX 5060 Ti 16 GB Palit Infinity 3

75

65

020406080

Black Myth Wukong (1080p High) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	71 Avg FPS, 58 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	74 Avg FPS, 62 1% Low FPS
RX 9060 XT 16 GB XFX	73 Avg FPS, 63 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	75 Avg FPS, 65 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

41

26

RTX 5060 Ti 8 GB Palit Dual

52

35

RX 9060 XT 16 GB XFX

49

39

RTX 5060 Ti 16 GB Palit Infinity 3

51

41

015304560

Cyberpunk 2077 (1080p RT Ultra) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	41 Avg FPS, 26 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	52 Avg FPS, 35 1% Low FPS
RX 9060 XT 16 GB XFX	49 Avg FPS, 39 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	51 Avg FPS, 41 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

82

66

RTX 5060 Ti 8 GB Palit Dual

90

73

RX 9060 XT 16 GB XFX

95

61

RTX 5060 Ti 16 GB Palit Infinity 3

90

71

0255075100

F1 24 (1080p Ultra high) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	82 Avg FPS, 66 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	90 Avg FPS, 73 1% Low FPS
RX 9060 XT 16 GB XFX	95 Avg FPS, 61 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	90 Avg FPS, 71 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

111

63

RTX 5060 Ti 8 GB Palit Dual

108

32

RX 9060 XT 16 GB XFX

106

65

RTX 5060 Ti 16 GB Palit Infinity 3

117

65

037.575112.5150

Homeworld 3 (1080p Epic) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	111 Avg FPS, 63 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	108 Avg FPS, 32 1% Low FPS
RX 9060 XT 16 GB XFX	106 Avg FPS, 65 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	117 Avg FPS, 65 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

93

64

RTX 5060 Ti 8 GB Palit Dual

101

65

RX 9060 XT 16 GB XFX

96

42

RTX 5060 Ti 16 GB Palit Infinity 3

98

42

037.575112.5150

Metro Exodus Enhanced Edition (1080p Ultra) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	93 Avg FPS, 64 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	101 Avg FPS, 65 1% Low FPS
RX 9060 XT 16 GB XFX	96 Avg FPS, 42 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	98 Avg FPS, 42 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

66

35

RTX 5060 Ti 8 GB Palit Dual

69

37

RX 9060 XT 16 GB XFX

73

37

RTX 5060 Ti 16 GB Palit Infinity 3

75

41

020406080

The Talos Principle 2 (1080p Ultra) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	66 Avg FPS, 35 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	69 Avg FPS, 37 1% Low FPS
RX 9060 XT 16 GB XFX	73 Avg FPS, 37 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	75 Avg FPS, 41 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

90

42

RTX 5060 Ti 8 GB Palit Dual

81

59

RX 9060 XT 16 GB XFX

94

72

RTX 5060 Ti 16 GB Palit Infinity 3

83

54

0255075100

Total War: Warhammer 3 (1080p Ultra) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	90 Avg FPS, 42 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	81 Avg FPS, 59 1% Low FPS
RX 9060 XT 16 GB XFX	94 Avg FPS, 72 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	83 Avg FPS, 54 1% Low FPS

A 41 fps average result at Ultra RT settings, compared to the 52 fps average of the Nvidia contender, is seriously off the pace. At this level, while both GPUs fail to meet the 60 fps "gosh that's smooth" minimum, an 11 frames per second difference is very significant.

That being said, I found the RTX 5060 Ti 8 GB to fly through Night City once the settings were dropped to Medium RT in my 8 GB vs 16 GB VRAM testing, so I looked back at my in-game results and… yeesh.

Cyberpunk 2077 - VRAM usage and performance

In-game testing, Medium RT settings

Cyberpunk 2077 (1080p Medium RT + Quality upscaling)Cyberpunk 2077 (1440p Medium RT + Quality upscaling)Cyberpunk 2077 (1440p Medium RT + Qual. upscaling + 2x FG)

Avg FPS

1% Low FPS

Avg. VRAM

RX 9060 XT 8 GB Asus Prime

72

51

7.68

RTX 5060 Ti 8 GB Palit Infinity 3

91

44

7.32

RX 9060 XT 16 GB XFX

85

67

9.46

RTX 5060 Ti 16 GB Palit Infinity 3

91

63

8.76

0255075100

Cyberpunk 2077 (1080p Medium RT + Quality upscaling) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	72 Avg FPS, 51 1% Low FPS, 7.68 Avg. VRAM
RTX 5060 Ti 8 GB Palit Infinity 3	91 Avg FPS, 44 1% Low FPS, 7.32 Avg. VRAM
RX 9060 XT 16 GB XFX	85 Avg FPS, 67 1% Low FPS, 9.46 Avg. VRAM
RTX 5060 Ti 16 GB Palit Infinity 3	91 Avg FPS, 63 1% Low FPS, 8.76 Avg. VRAM

Avg FPS

1% Low FPS

Avg. VRAM

RX 9060 XT 8 GB Asus Prime

45

27

7.6

RTX 5060 Ti 8 GB Palit Infinity 3

58

37

7.32

RX 9060 XT 16 GB XFX

57

47

11.15

RTX 5060 Ti 16 GB Palit Infinity 3

58

45

10.01

015304560

Cyberpunk 2077 (1440p Medium RT + Quality upscaling) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	45 Avg FPS, 27 1% Low FPS, 7.6 Avg. VRAM
RTX 5060 Ti 8 GB Palit Infinity 3	58 Avg FPS, 37 1% Low FPS, 7.32 Avg. VRAM
RX 9060 XT 16 GB XFX	57 Avg FPS, 47 1% Low FPS, 11.15 Avg. VRAM
RTX 5060 Ti 16 GB Palit Infinity 3	58 Avg FPS, 45 1% Low FPS, 10.01 Avg. VRAM

Avg FPS

1% Low FPS

Avg. VRAM

RX 9060 XT 8 GB Asus Prime

86

45

7.68

RTX 5060 Ti 8 GB Palit Infinity 3

94

43

7.33

RX 9060 XT 16 GB XFX

103

63

11.84

RTX 5060 Ti 16 GB Palit Infinity 3

99

74

10.52

037.575112.5150

Cyberpunk 2077 (1440p Medium RT + Qual. upscaling + 2x FG) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	86 Avg FPS, 45 1% Low FPS, 7.68 Avg. VRAM
RTX 5060 Ti 8 GB Palit Infinity 3	94 Avg FPS, 43 1% Low FPS, 7.33 Avg. VRAM
RX 9060 XT 16 GB XFX	103 Avg FPS, 63 1% Low FPS, 11.84 Avg. VRAM
RTX 5060 Ti 16 GB Palit Infinity 3	99 Avg FPS, 74 1% Low FPS, 10.52 Avg. VRAM

The RX 9060 XT 8 GB is a full 19 frames off the pace at 1080p in Cyberpunk's open world, even with the settings turned down—although it does manage a significantly higher 1% low result.

At 1440p, however, the AMD card is still 13 frames behind the RTX 5060 Ti 8 GB, and the 1 % lows plummet, too. The RX 9060 XT 8 GB gains some ground back with upscaling and frame generation enabled, though, which is some small comfort.

Going back to my 1080p benchmarks for a second, it's clear that it's not just Cyberpunk 2077 that gives this card some trouble compared to its competition. It's behind the pack in all but two of my tests, although with much less of a margin.

It does manage to beat the RTX 5060 Ti 8 GB by a full nine frames in Total War: Warhammer 3, though, and even manages to surge ahead of the 16 GB version.

1440p gaming benchmarks | RTX 9060 XT 8 GB

Black Myth Wukong (1440p High)Cyberpunk 2077 (1440p RT Ultra)F1 24 (1440p Ultra High)Homeworld 3 (1440p Epic)Metro Exodus Enhanced Edition (1440p Ultra)The Talos Principle 2 (1440p Ultra)Total War: Warhammer 3 (1440p Ultra)

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

48

43

RTX 5060 Ti 8 GB Palit Dual

50

43

RX 9060 XT 16 GB XFX

48

44

RTX 5060 Ti 16 GB Palit Infinity 3

51

45

015304560

Black Myth Wukong (1440p High) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	48 Avg FPS, 43 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	50 Avg FPS, 43 1% Low FPS
RX 9060 XT 16 GB XFX	48 Avg FPS, 44 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	51 Avg FPS, 45 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

26

17

RTX 5060 Ti 8 GB Palit Dual

33

25

RX 9060 XT 16 GB XFX

30

24

RTX 5060 Ti 16 GB Palit Infinity 3

32

26

010203040

Cyberpunk 2077 (1440p RT Ultra) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	26 Avg FPS, 17 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	33 Avg FPS, 25 1% Low FPS
RX 9060 XT 16 GB XFX	30 Avg FPS, 24 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	32 Avg FPS, 26 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

51

41

RTX 5060 Ti 8 GB Palit Dual

59

47

RX 9060 XT 16 GB XFX

60

46

RTX 5060 Ti 16 GB Palit Infinity 3

56

45

015304560

F1 24 (1440p Ultra High) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	51 Avg FPS, 41 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	59 Avg FPS, 47 1% Low FPS
RX 9060 XT 16 GB XFX	60 Avg FPS, 46 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	56 Avg FPS, 45 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

75

58

RTX 5060 Ti 8 GB Palit Dual

87

42

RX 9060 XT 16 GB XFX

86

65

RTX 5060 Ti 16 GB Palit Infinity 3

97

64

0255075100

Homeworld 3 (1440p Epic) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	75 Avg FPS, 58 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	87 Avg FPS, 42 1% Low FPS
RX 9060 XT 16 GB XFX	86 Avg FPS, 65 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	97 Avg FPS, 64 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

68

49

RTX 5060 Ti 8 GB Palit Dual

75

53

RX 9060 XT 16 GB XFX

71

34

RTX 5060 Ti 16 GB Palit Infinity 3

72

34

020406080

Metro Exodus Enhanced Edition (1440p Ultra) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	68 Avg FPS, 49 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	75 Avg FPS, 53 1% Low FPS
RX 9060 XT 16 GB XFX	71 Avg FPS, 34 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	72 Avg FPS, 34 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

47

27

RTX 5060 Ti 8 GB Palit Dual

44

30

RX 9060 XT 16 GB XFX

49

35

RTX 5060 Ti 16 GB Palit Infinity 3

51

32

015304560

The Talos Principle 2 (1440p Ultra) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	47 Avg FPS, 27 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	44 Avg FPS, 30 1% Low FPS
RX 9060 XT 16 GB XFX	49 Avg FPS, 35 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	51 Avg FPS, 32 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

60

44

RTX 5060 Ti 8 GB Palit Dual

57

40

RX 9060 XT 16 GB XFX

62

45

RTX 5060 Ti 16 GB Palit Infinity 3

60

44

020406080

Total War: Warhammer 3 (1440p Ultra) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	60 Avg FPS, 44 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	57 Avg FPS, 40 1% Low FPS
RX 9060 XT 16 GB XFX	62 Avg FPS, 45 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	60 Avg FPS, 44 1% Low FPS

Switching things up to 1440p, the RX 9060 XT 8 GB drops behind once again in all but two of my tests, although the margins are a fair bit closer on average. Cyberpunk 2077 proves to be this card's Achilles' heel, which suggests the ray tracing demands may be a little too much for its 8 GB loadout.

It's also worth noting the difference between the 8 GB GPU and its 16 GB-equipped bigger brother. This difference can be quite pronounced in certain games, particularly in Homeworld 3, where the RX 9060 XT 8 GB posts a 75 fps average frame rate result to the 16 GB version's 97 frames.

Again, I have to remind myself that this 8 GB card is by far the cheapest of all my contenders. But with the $40-$50 more expensive RTX 5060 Ti 8 GB ranging ahead on average, sometimes by significant amounts, I do have to wonder if that saving is really worth it.

Real-world gaming performance | RX 9060 XT 8 GB

Black Myth Wukong (1440p High + Quality upscaling)Cyberpunk 2077 (1440p RT Ultra + Quality upscaling)F1 24 (1440p Ultra high + Quality upscaling, 2x FG)Homeworld 3 (1440p Epic + Quality upscaling)The Talos Principle 2 (1440p Ultra + Quality upscaling, 2x FG)

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

76

68

RTX 5060 Ti 8 GB Palit Dual

79

68

RX 9060 XT 16 GB XFX

78

69

RTX 5060 Ti 16 GB Palit Infinity 3

86

72

0255075100

Black Myth Wukong (1440p High + Quality upscaling) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	76 Avg FPS, 68 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	79 Avg FPS, 68 1% Low FPS
RX 9060 XT 16 GB XFX	78 Avg FPS, 69 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	86 Avg FPS, 72 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

90

60

RTX 5060 Ti 8 GB Palit Dual

93

66

RX 9060 XT 16 GB XFX

102

73

RTX 5060 Ti 16 GB Palit Infinity 3

96

79

037.575112.5150

Cyberpunk 2077 (1440p RT Ultra + Quality upscaling) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	90 Avg FPS, 60 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	93 Avg FPS, 66 1% Low FPS
RX 9060 XT 16 GB XFX	102 Avg FPS, 73 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	96 Avg FPS, 79 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

146

122

RTX 5060 Ti 8 GB Palit Dual

131

104

RX 9060 XT 16 GB XFX

190

113

RTX 5060 Ti 16 GB Palit Infinity 3

136

89

050100150200

F1 24 (1440p Ultra high + Quality upscaling, 2x FG) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	146 Avg FPS, 122 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	131 Avg FPS, 104 1% Low FPS
RX 9060 XT 16 GB XFX	190 Avg FPS, 113 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	136 Avg FPS, 89 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

102

51

RTX 5060 Ti 8 GB Palit Dual

113

46

RX 9060 XT 16 GB XFX

110

62

RTX 5060 Ti 16 GB Palit Infinity 3

118

63

037.575112.5150

Homeworld 3 (1440p Epic + Quality upscaling) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	102 Avg FPS, 51 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	113 Avg FPS, 46 1% Low FPS
RX 9060 XT 16 GB XFX	110 Avg FPS, 62 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	118 Avg FPS, 63 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

127

78

RTX 5060 Ti 8 GB Palit Dual

103

71

RX 9060 XT 16 GB XFX

135

59

RTX 5060 Ti 16 GB Palit Infinity 3

115

41

037.575112.5150

The Talos Principle 2 (1440p Ultra + Quality upscaling, 2x FG) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	127 Avg FPS, 78 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	103 Avg FPS, 71 1% Low FPS
RX 9060 XT 16 GB XFX	135 Avg FPS, 59 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	115 Avg FPS, 41 1% Low FPS

Thankfully, my upscaling and frame generation testing gives the RX 9060 XT 8 GB something to celebrate. This tightens up the difference between the two 8 GB cards on average, and even leads to some eyebrow-raising figures.

Take F1 24, for example. We've observed this game posting astonishing frame rate scores with AMD's FSR tech enabled before (check out the massive 292 fps result in our RX 9070 XT review, for example), and that tracks with the 146 fps average here.

If you really want to see an AMD GPU fly, though, check out the RX 9060 XT 16 GB further down the chart. 190 fps?!? That beats the RTX 5060 Ti 16 GB by a full 54 frames. And the 8 GB card by *checks calculator* 44 fps. What a difference an extra 8 GB of VRAM makes, ey?

And check out those The Talos Principle 2 figures, while we're at it. It also seems to prefer FSR over DLSS, which suggests that F1 24 might not be the only game that likes what AMD brings to the table.

Still, while the upscaling results muddy the waters a little, it's clear that the little RX 9060 XT 8 GB is bested in most of my tests by its more expensive competition. And bringing some more real-world, in-game testing into the mix makes that clearer.

The Last of Us Part 1 - VRAM usage and performance

In-game testing, Ultra settings

The Last of Us Part 1 (1080p Ultra + Quality upscaling)The Last of Us Part 1 (1440p Ultra + Quality upscaling)The Last of Us Part 1 (1440p Ultra+ Quality upscaling + 2x FG)

Avg FPS

1% Low FPS

Avg. VRAM

RX 9060 XT 8 GB Asus Prime

79

52

6.61

RTX 5060 Ti 8 GB Palit Dual

89

48

6.92

RX 9060XT 16 GB XFX

113

93

11.47

RTX 5060 Ti 16 GB Palit Infinity 3

113

92

9.84

037.575112.5150

The Last of Us Part 1 (1080p Ultra + Quality upscaling) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	79 Avg FPS, 52 1% Low FPS, 6.61 Avg. VRAM
RTX 5060 Ti 8 GB Palit Dual	89 Avg FPS, 48 1% Low FPS, 6.92 Avg. VRAM
RX 9060XT 16 GB XFX	113 Avg FPS, 93 1% Low FPS, 11.47 Avg. VRAM
RTX 5060 Ti 16 GB Palit Infinity 3	113 Avg FPS, 92 1% Low FPS, 9.84 Avg. VRAM

Avg FPS

1% Low FPS

Avg. VRAM

RX 9060 XT 8 GB Asus Prime

60

42

6.71

RTX 5060 Ti 8 GB Palit Dual

79

44

6.98

RX 9060XT 16 GB XFX

89

72

12.02

RTX 5060 Ti 16 GB Palit Infinity 3

92

73

10.39

0255075100

The Last of Us Part 1 (1440p Ultra + Quality upscaling) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	60 Avg FPS, 42 1% Low FPS, 6.71 Avg. VRAM
RTX 5060 Ti 8 GB Palit Dual	79 Avg FPS, 44 1% Low FPS, 6.98 Avg. VRAM
RX 9060XT 16 GB XFX	89 Avg FPS, 72 1% Low FPS, 12.02 Avg. VRAM
RTX 5060 Ti 16 GB Palit Infinity 3	92 Avg FPS, 73 1% Low FPS, 10.39 Avg. VRAM

Avg FPS

1% Low FPS

Avg. VRAM

RX 9060 XT 8 GB Asus Prime

112

61

6.74

RTX 5060 Ti 8 GB Palit Dual

132

61

7.21

RX 9060XT 16 GB XFX

149

104

12.12

RTX 5060 Ti 16 GB Palit Infinity 3

147

78

11.09

037.575112.5150

The Last of Us Part 1 (1440p Ultra+ Quality upscaling + 2x FG) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	112 Avg FPS, 61 1% Low FPS, 6.74 Avg. VRAM
RTX 5060 Ti 8 GB Palit Dual	132 Avg FPS, 61 1% Low FPS, 7.21 Avg. VRAM
RX 9060XT 16 GB XFX	149 Avg FPS, 104 1% Low FPS, 12.12 Avg. VRAM
RTX 5060 Ti 16 GB Palit Infinity 3	147 Avg FPS, 78 1% Low FPS, 11.09 Avg. VRAM

I've picked The Last of Us Part 1 for a reason. Despite many, many patches since its release, it's still known as something of a VRAM hog, and can soon trip up lower-spec cards with the settings turned up. And while the RX 9060 XT 8 GB still manages smooth frame rates in this game at Ultra settings, it really does show the difference between this GPU and my other tested cards.

The AMD card is 20 frames behind the RTX 5060 Ti 8 GB at 1080p on average, and 35 fps behind the RX 9060 XT 16 GB. Switch the resolution to 1440p, and it just about squeaks a 60 fps average while the other GPUs are ranging off into the high 70s and low 90s.

Even frame generation can't quite make up the difference for the RX 9060 XT 8 GB, which again ends up far behind all our other cards.

4K gaming benchmarks | RX 9060 XT 8 GB

Black Myth Wukong (4K High)Cyberpunk 2077 (4K RT Ultra)F1 24 (4K Ultra High)Homeworld 3 (4K Epic)Metro Exodus Enhanced Edition (4K Ultra)The Talos Principle 2 (4K Ultra)Total War: Warhammer 3 (4K Ultra)

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

25

22

RTX 5060 Ti 8 GB Palit Dual

26

21

RX 9060 XT 16 GB XFX

25

22

RTX 5060 Ti 16 GB Palit Infinity 3

28

25

07.51522.530

Black Myth Wukong (4K High) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	25 Avg FPS, 22 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	26 Avg FPS, 21 1% Low FPS
RX 9060 XT 16 GB XFX	25 Avg FPS, 22 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	28 Avg FPS, 25 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

8

5

RTX 5060 Ti 8 GB Palit Dual

8

4

RX 9060 XT 16 GB XFX

14

11

RTX 5060 Ti 16 GB Palit Infinity 3

15

12

03.757.511.2515

Cyberpunk 2077 (4K RT Ultra) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	8 Avg FPS, 5 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	8 Avg FPS, 4 1% Low FPS
RX 9060 XT 16 GB XFX	14 Avg FPS, 11 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	15 Avg FPS, 12 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

24

20

RTX 5060 Ti 8 GB Palit Dual

26

21

RX 9060 XT 16 GB XFX

30

22

RTX 5060 Ti 16 GB Palit Infinity 3

27

21

07.51522.530

F1 24 (4K Ultra High) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	24 Avg FPS, 20 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	26 Avg FPS, 21 1% Low FPS
RX 9060 XT 16 GB XFX	30 Avg FPS, 22 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	27 Avg FPS, 21 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

38

25

RTX 5060 Ti 8 GB Palit Dual

49

33

RX 9060 XT 16 GB XFX

46

41

RTX 5060 Ti 16 GB Palit Infinity 3

58

50

015304560

Homeworld 3 (4K Epic) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	38 Avg FPS, 25 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	49 Avg FPS, 33 1% Low FPS
RX 9060 XT 16 GB XFX	46 Avg FPS, 41 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	58 Avg FPS, 50 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

36

27

RTX 5060 Ti 8 GB Palit Dual

43

33

RX 9060 XT 16 GB XFX

41

21

RTX 5060 Ti 16 GB Palit Infinity 3

42

20

015304560

Metro Exodus Enhanced Edition (4K Ultra) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	36 Avg FPS, 27 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	43 Avg FPS, 33 1% Low FPS
RX 9060 XT 16 GB XFX	41 Avg FPS, 21 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	42 Avg FPS, 20 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

22

8

RTX 5060 Ti 8 GB Palit Dual

20

10

RX 9060 XT 16 GB XFX

25

19

RTX 5060 Ti 16 GB Palit Infinity 3

27

18

07.51522.530

The Talos Principle 2 (4K Ultra) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	22 Avg FPS, 8 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	20 Avg FPS, 10 1% Low FPS
RX 9060 XT 16 GB XFX	25 Avg FPS, 19 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	27 Avg FPS, 18 1% Low FPS

Avg FPS

1% Low FPS

RX 9060 XT 8 GB Asus Prime

30

19

RTX 5060 Ti 8 GB Palit Dual

32

21

RX 9060 XT 16 GB XFX

30

20

RTX 5060 Ti 16 GB Palit Infinity 3

33

22

010203040

Total War: Warhammer 3 (4K Ultra) Data

Product	Value
RX 9060 XT 8 GB Asus Prime	30 Avg FPS, 19 1% Low FPS
RTX 5060 Ti 8 GB Palit Dual	32 Avg FPS, 21 1% Low FPS
RX 9060 XT 16 GB XFX	30 Avg FPS, 20 1% Low FPS
RTX 5060 Ti 16 GB Palit Infinity 3	33 Avg FPS, 22 1% Low FPS

If you need some further proof that 16 GB of VRAM can sometimes make a difference, you only have to look at the 4K results. No-one is suggesting that any of these GPUs are built for 4K gaming, nor do they need to be. But the averages are still interesting as an example of how these graphics cards can behave when pushed to their absolute limits.

Both our 8 GB contenders suffer more than their 16 GB equivalents overall, although the RTX 5060 Ti 8 GB still manages to surprise on occasion. The little RX 9060 XT 8 GB does beat it by two frames in The Talos Principle 2, though, and also manages to match its 16 GB variant in two of my tests, albeit at crunchy frame rates. Small wins, and all that.

PC Gamer test platform
Supplied by Cyberpower | MSI

CPU: AMD Ryzen 7 9800X3D | Motherboard: MSI MPG X870E Edge Ti WiFi | RAM: Kingston Fury Beast RGB 32 GB (2 x 16 GB) @ 6,000 MT/s | Cooler: MAG CoreLiquid i360 White | SSD: Spatium M480 Pro 2 TB | PSU: MPG A1000GS PCIe 5 | Case: MAG Pano 100R White

RX 9060 XT 8 GB - Thermals & Power

It should be no surprise to learn that this particular Asus Prime variant is a very cool running GPU, given that it's got a hilariously-oversized cooler for a budget graphics card. What's interesting, though, is that the extra size (and fan) results in a 6 °C lower temperature than my 16 GB sample, which has a small, dual-fan, dual-slot cooler.

Gaming temperatures | RX 9060 XT 8 GB

Thermals

Peak temperature (°C)

Average temperature (°C)

RX 9060 XT 8 GB Asus Prime

51

48

RTX 5060 Ti 8 GB Palit Dual

68

64

RX 9060 XT 16 GB XFX

57

48

RTX 5060 Ti 16 GB Palit Infinity 3

67

65

020406080

Thermals Data

Product	Value
RX 9060 XT 8 GB Asus Prime	51 Peak temperature (°C), 48 Average temperature (°C)
RTX 5060 Ti 8 GB Palit Dual	68 Peak temperature (°C), 64 Average temperature (°C)
RX 9060 XT 16 GB XFX	57 Peak temperature (°C), 48 Average temperature (°C)
RTX 5060 Ti 16 GB Palit Infinity 3	67 Peak temperature (°C), 65 Average temperature (°C)

With a 51 °C peak result and a 48 °C average, the Asus Prime RX 9060 XT 8 GB OC Edition is one of the most chilled out graphics cards I've ever tested. You're trading these temps for some considerable bulk, though, and it's more common to see dual-fan coolers on most examples.

Both our RTX 9060 XT samples run much cooler than the equivalent Nvidia cards, which likely means that most dual-fan versions should be good candidates for small form factor builds. Just not this Asus variant. You'd have some trouble cramming it in.

Power benchmarks | RX 9060 XT 8 GB

Higher is better

Performance per watt

4K frames per J

1080p frame per J

RX 9060 XT 8 GB Asus Prime

0.22

0.56

RTX 5060 Ti 8 GB Palit Dual

0.23

0.56

RX 9060 XT 16 GB XFX

0.28

0.65

RTX 5060 Ti 16 GB Palit Infinity 3

0.23

0.62

00.20.40.60.8

Performance per watt Data

Product	Value
RX 9060 XT 8 GB Asus Prime	0.22 4K frames per J, 0.56 1080p frame per J
RTX 5060 Ti 8 GB Palit Dual	0.23 4K frames per J, 0.56 1080p frame per J
RX 9060 XT 16 GB XFX	0.28 4K frames per J, 0.65 1080p frame per J
RTX 5060 Ti 16 GB Palit Infinity 3	0.23 4K frames per J, 0.62 1080p frame per J

The RX 9060 XT 8 GB has a 150 W TGP, but just like the 16 GB card (and many others), it will occasionally peak significantly beyond that under heavy load. I recorded a maximum peak of 175 W and an average power use of 164 W under repeated runs of the Metro Exodus Enhanced Edition benchmark, using this Asus Prime example.

Averaging that out by frame rate results in a score of 0.56 frames per joule at 1080p, and 0.23 frames per joule at 4K. That's almost identical to the RTX 5060 Ti 8 GB Palit Dual sample I tested, once the performance gain is factored into its 180 W power limit.

RX 9060 XT 8 GB - Sound

Given that my review sample has an aforementioned massive cooler, it's remarkable how little it has to push to keep the RX 9060 XT 8 GB chilled. Don't be fooled by those three large fans—they make very little noise, even when this card's pushed to its limits.

That being said, the twin-fan cooler on the Palit Dual RTX 5060 Ti 8 GB is even quieter. Whichever way you look at it, though, this is one hushed-up GPU.

RTX 9060 XT 8 GB - Software

I've got a fair bit of experience with AMD Adrenaline, and can confirm that it's pretty good these days at allowing you to get involved with your graphics card. It provides easy access to overclocking tools and driver updates, a handy homepage listing of your most recently played games and the average recorded frame rate, and lots of extra features that can be enabled on a per-game basis, like AMD Fluid Motion Frames.

While the Nvidia App looks quite different, the functionality between the two is pretty comparable, so you won't be missing out on anything vital if you opt for team red over team green.

In terms of software provided by Asus, well, you could always download Asus GPU Tweak III, an overclocking tool that I'm also rather fond of. It's not pretty, but GPU Tweak III is a straightforward way of pushing most modern graphics cards to their absolute limits.

It's arguably a bit redundant given that this GPU can be overclocked directly within AMD Adrenaline, but hey, extras are extras. Although it's not really an extra, as anyone can use it for free without purchasing an Asus card. Anyway, software and driver wise? I've had zero issues with either, and the RX 9060 XT 8 GB certainly isn't left wanting as a result.

RX 9060 XT 8 GB - Value

You can make the argument that, given its very reasonable price, the RX 9060 XT 8 GB is still a decent budget graphics card for gaming. And that's true. But, as the RTX 5060 Ti 8 GB is capable of leaping ahead when the VRAM constraints become tight, it's tough to argue that the Nvidia card isn't worth the extra cash.

If you could find an RX 9060 XT 8 GB for its $299 MSRP, I'd be much more inclined to give it a higher score. However, we keep a beady eye on graphics card pricing all year long, and the cheapest version we can find at the time of writing is $350. That's still fairly reasonable given the performance (and in today's memory crisis-influenced market), but when the RTX 5060 Ti 8 GB can be found for $40-$50 more, I know which I'd be picking.

The Nvidia card beats the little AMD in most of my benchmarks, and in VRAM-constrained scenarios, it's much better at holding its drink. And while I've got no crystal ball available to look into the future, I'd say it was a safe bet that if an 8 GB VRAM limit can cause significant performance loss now, it's unlikely to get any better as time goes on.

Futureproofing is a dirty word in my business, but it's still prudent to buy a graphics card with a decent degree of headroom now to prevent having to upgrade it sooner than you'd like. And while the RTX 5060 Ti 8 GB has some of the same issues as the 8 GB AMD card (lower overall performance compared to its 16 GB version), the fact that the drop between the two is lesser makes the Nvidia GPU the better buy.

That being said, if money's super tight and you can't stretch to the Nvidia card (or the RX 9060 XT 16 GB), then the RX 9060 XT 8 GB isn't too bad. It still posts reasonable frame rates in many of my tests, and it's encouraging to see that the addition of upscaling and frame generation can tighten up the performance gap in many scenarios.

Still, I can't help but be slightly underwhelmed by AMD's 8 GB effort overall. Even this Asus Prime Radeon RX 9060 XT 8 GB OC Edition variant is a tad disappointing, especially when you consider it's currently $500. For that price? No way.

A lesser-cooled card for $350? Sure, in a pinch. But whenever I played a game where the frame rate crunched, I think I'd be wondering whether the RTX 5060 Ti 8 GB would have gained me a few frames—or perhaps even a lot more in certain scenarios. And, it must be said, whether I should have put an extra $50 towards a different GPU instead.

Have you ever looked at a piece of tech and had to fight the urge to crunch it like rock candy? Well, take a look at the recently announced wireless ZenMouse MD202 from Asus and get back to me.

Personally, I blame the mouse's 'Ceraluminum' top cover for making it look like a tasty snack in every sense. Thankfully, the ZenMouse MD202 is designed to survive much more than any ill-advised snacking, with the Ceraluminum finish offering neutral office aesthetics alongside "exceptional wear-, scratch-, and shock-resistance" as per the press release.

This isn't Asus' first bit of kit to enjoy a ceramized aluminium finish, with the company unveiling the same hard-wearing top coat on a number of its Zenbooks back in 2024. However, the ZenMouse MD202 hasn't really been built with a gamer audience top of mind, though it does still offer a number of features gamers will find especially tasty.

For a start, you have the option of either a Bluetooth (yuck) or a 2.4 GHz wireless connection (yay). The mouse's tactile micro-switches are also described as "gaming-grade" on account of being able to withstand up to 50 million clicks.

"Full-perimeter PTFE mouse feet" are also meant to ensure smooth moves from this mouse, though that's framed around 'enhancing workflows' rather than rebalancing your KDA. At 85 g, it's also not really the most lightweight mouse for gaming.

The ZenMouse MD202 offers adjustable sensitivity levels ranging from 800 to 4200 DPI, though I'm sure I don't need to highlight how that's actually pretty low compared to what's offered by the best gaming mice out there.

The actual big win in my book as a wireless mouse is its battery life. For one thing, the ZenMouse MD202 enjoys fast charging via USB-C, slurping up 16 hours' worth of runtime juice from just one minute of charging.

Asus otherwise resists attributing a straightforward hour count to a full charge, instead saying that the rechargeable battery will give you up to seven months of on-time, once fully powered. Asus has yet to announce a release date or MSRP for the ZenMouse MD202, though with a battery and surface material like that, you can bet this won't be a budget option for anyone.

A fresh report from Charlie Demerjian's SemiAccurate pegged Nvidia to be negotiating to buy a "large PC oriented company". Though the GPU company has since denied it in later reporting by Bloomberg.

The paywalled report from SemiAccurate suggests that the purchase would "fundamentally change the computing landscape". The report from the respected outlet was enough to see a surge in share prices for a range of companies on April 13. While we can't be sure these are entirely based on the report, HP, Dell, Asus and Lenovo all saw similar rises that day—as much as 4%.

Though the markets have since cooled off.

Bloomberg received a denial of the claims in no uncertain terms later that same day: "The media report is false; Nvidia is not engaged in discussions to acquire any PC maker," a spokesperson for Nvidia told Bloomberg.

Dermerjian is well regarded for his reporting in the technology industry, and cites that the website is "dead serious" about the claims. He also cites a similar story when Elon Musk sought to purchase Intel, and suggests this since evolved into Intel's heavy involvement in Musk's Terafab project. The Terafab project is intended to boost fab output for AI chips in the future.

While it's hard to get a grasp on whether there's any suggestion of a future purchase by Nvidia here, the company has made it clear that it wants a larger slice of the pie in regards to both datacenter and client computing. With an Arm chip in hand—Nvidia had once tried to purchase Arm itself but was blocked by regulators—it is reportedly taking aim at client computing with the N1X chip. N1X looks to have a pretty powerful processor behind it, alongside a GPU with 6,144 CUDA cores—equivalent to an RTX 5070.

Nvidia has also invested in Intel. This partnership will see Intel working alongside the GPU maker on yet to be confirmed x86 collaboration and datacenter collaboration. So whether Nvidia wants to buy a PC manufacturer or not, and currently it looks to be more the latter than anything, it appears it will be working a lot closer with some of them in the near-future anyways.

If you frequently haunt the usual social media channels for PC enthusiasts, you'll notice that nary a day goes by without someone posting an image of their expensive graphics card, replete with a melting power connector. Various efforts to combat the problem have yet to completely nix the issue, but Asus reckons it now has the solution: one seriously beefy power cable.

Called the ROG Equalizer, it looks just like any other premium 12VHPWR/12V-2x6 power cable, but there's more to it than meets the eye. Asus says that the new design "ensures balanced current delivery from PSU to ROG Equalizer" so that you shouldn't get situations where a single 12 V pin has to deal with the full 575 W of an RTX 5090, for example.

But for situations where things do go awry, each power cable is rated to 17 A of current draw (compared to 9.2 A as per the 12V-2x6 specification), and there's an additional cable integrated into the Equalizer, that you can plug into a little socket in Asus' ROG Thor 3 and ROG Strix PSUs.

At some point in the near future, GPU Tweak III (a software tool by Asus that works a bit like MSI Afterburner) will be updated to have a 'ROG Equaliser mode', which should give you a real-time monitor of current draw in each 12V-2x6 pin.

Asus isn't the only company to offer such a feature, of course, as MSI and ASRock have their own solutions, though this seems to be the first instance of creating a more capable cable that I've seen so far.

While there's no indication as to the ROG Equalizer's price tag, the good news is that you don't need to own an Asus megabucks PSU to use it (though you will be able to get the cable at a discounted price, if you do).

Since the ROG Equalizer uses standard 12V-2x6 connectors at both ends of the cable, it will work in any PSU that sports a similar socket.

In Asus' own words: "[It has] no specific model requirements. The ROG Equalizer is bundled with the 2026 ASUS ROG Thor III and ASUS ROG Strix Platinum power supplies, and is also compatible with power supplies (ATX3.1 with native 12V-2x6 connector) from all leading manufacturers."

I can foresee many an RTX 5090 owner eyeing a purchase right now.

Crimson Desert is one of those games that seems to be able to wrangle together all kinds of disparate gamers to attempt to master the bizarre behemoth of a multi-tool RPG. So it's a little unfortunate that it has had far from perfect hardware support. We witnessed the game's lack of Intel Arc support ourselves, and now Pearl Abyss and Asus are saying ROG Xbox Ally X devices might need to roll back their drivers to launch the game.

In its list of known issues, the game developer and publisher, Pearl Abyss, says: "We have identified an issue where Crimson Desert is not launching properly on ROG Xbox Ally X running on the latest driver version. Microsoft and AMD are currently looking into the issue. Until the issue is resolved, we recommend installing and using the driver version immediately prior to the latest one, available on the Asus website."

Asus reiterates the same and provides a download link that, uhh, doesn't work. If you want to download the driver to roll it back, go to the ROG Xbox Ally X driver and tools page, scroll down to Graphics, click See All Downloads, and download the driver Version V32.0.21025.27003.

It's never a nice feeling to have to roll back your drivers, but Pearl Abyss assures us that this is only until the issue is resolved. And if a driver rollback really does work to get the game launching on the Microsoft-Asus handheld, there's good reason to think it can get up and running on the latest drivers with a little work, too.

Hopefully, these compatibility issues will be resolved quickly. And ditto the Intel Arc ones, given after the initial issues were spotted, Pearl Abyss said it will work on bringing support for the GPUs to the game, and Intel told us it "remains ready to assist Pearl Abyss however we can."

Even if the game does work on your hardware, although general performance seems to be smooth, plenty of people have reported that the game is a little graphically glitchy, blurry, and so on. In which case, I suppose it's good news that the devs are actively working on improving things.

No, the ROG Kithara is not the gaming headset to solve the cost of living crisis. It is a joy to listen to your games and music with, though, and a genuine curiosity in a crowded market. You don't need one, and neither do I, but let's sit awhile anyway and enjoy what ROG and Hifiman have built.

This one's all about the drivers. Vast, 100 mm planar magnetic ones, designed and built in partnership with New York-based audio specialists, Hifiman. Every other element's designed to let those drivers do their thing. An open-backed design that doesn't care one bit about the fact you can hear the tumble drier from your gaming room, or that you're sitting in a quiet coach. A frequency response range broader than a politician's promise. Y'know, audiophile stuff.

It's a wired model, featuring a number of different adapters that make it surprisingly compatible across PC, Xbox and PlayStation 5. You can plug it into a DAC if you like. And who are we kidding, of course you like. You're considering dropping $300 on a pair of headphones in this economy.

I'm inclined to think of the Asus ROG Kithara as a pair of headphones, rather than a gaming headset, because if you detach the super cardioid mic the only concession to the world of gaming is the ROG logo itself. None of the usual comforts of a gaming headset are present here: no physical volume control, no RGB zones, no digital surround. Certainly no chat mix adjustment, goodness no.

No, it's more like the Kithara is visiting us gamers from the Manhattan loft apartment of a wealthy hifi enthusiast to see what all this ‘Counter-Strike' stuff is about. Its visual cues are all reminiscent of Hifiman's headphone inventory, from the trademark elongated ovals of the earcups you'd see on the Ananda to the suspended headband beneath an aluminium frame, like an Arya.

(If you were wondering, ‘Kithara' is the ancient Greek stringed instrument from which we get the word ‘guitar'. Hang around in the games industry long enough and you're bound to pick up a bit of ancient Greek. If Kratos' adventures don't rub off on you, the peripheral market's fondness for an obscure product name inspired by classical mythology surely will.)

The sound is, of course, extraordinary. It's a far flatter response than most gaming headsets are tuned to produce, particularly at the low end, and if your ears have been addled by the bass response of Razer headsets circa 2011 you might initially perceive a lack of ‘oomph'. But that would be incorrect.

Instead, the Kithara's drivers give all frequencies the room they need to articulate clearly and distinctly, leading to a far richer sound than you get with a $100 headset that's had its mids scooped like a Metallica album. And if you were in any doubt about that EQ response, there's a bit of paperwork inside the box which shows the factory test results on a graph. It's marketing fluff, obviously, but having your eyes met by a ‘sound signature certificate' as you open the box is a nice touch.

The stellar sound reproduction is another reason I keep thinking about these as headphones and not a gaming headset. Because truthfully, I've had the most enjoyable experiences testing the Kithara with music, not games. There are a few games where the wide stereo spread and flat, spacious sound reproduction are genuinely useful, like Counter-Strike 2 and PUBG, where the Kithara picks up footsteps and far-off gunfire with artisanal care. And cinematic games like Crimson Desert whose soundtrack sounds especially primed to tug at your forearm hairs when it's heard through these magnetic planar drivers. They sound great, these games, but you're experiencing them without the usual comforts like wireless operation, chat mix adjustment or an easily locatable mic mute (this one sits inline on the mic cable).

There are no such downsides to enjoying its mellifluous tones in a musical setting, though. It's when I'm playing Yazz Ahmed albums with the sound up that the Kithara feels most at home, doing what it was designed to.

I'm not completely sold on the comfort. While the build quality, material choices and finishes are all absolutely magnificent and easily befitting of the price point, the fact is that this is a heavy 420 g unit and the earcups are big enough to park your car in. The suspended headband design does its best to distribute as much of that weight as possible over a wide area using a broad protein leather contact surface, but the headset still relies on a lot of clamping force to stay put. My head's on the smaller side, and I found that after about an hour I was, if not uncomfortable, then certainly aware I was wearing a headset.

Worse still, it was pointed out on several work calls that I looked… a bit odd wearing it. With the suspended headband pushed down to its lowest point to fit my smaller head, I'm left with about three inches of air between my head and the headset's aluminium upper band. Coupled with the oversized earcups, this was a source of fun to numerous colleagues. Fortunately/unfortunately, I was able to hear that derision with the kind of crystal clarity only a wired, open-backed model with 100 mm drivers can deliver. But I won't subtract marks for that.

Mic performance is better than Zoom fashion cred though. ROG's opted for a super cardioid polar pattern on this detachable mic, which allows you to point precisely at what you want to be picked up. (In most cases, that'll be your mouth.) That serves as a viable alternative to the digital noise suppression that most gaming headset mics employ, effectively ducking out the keyboard thwocks from your Discord chat.

This headset exists in a very rare sub-category, and it achieves almost everything it sets out to. It reminds me of Audio-Technica's excellent wired ATH-G1, only dialled up to ridiculous. In order to really establish itself as a gaming headset, it needs more features gamers find useful like a comprehensive and ergonomic physical controls layout. But you get the feeling that isn't the point of the exercise. This is aspirational kit, for people who can afford not to worry too much about the price tag, or the value for money. It's a branding exercise that establishes ROG's chops at delivering audiophile-grade sound, albeit in partnership with Hifiman. It does that. That doesn't mean you need it. But it does mean it's a ludicrously enjoyable listening experience.

After the US Federal Communications Commission (FCC) put consumer-grade routers produced in foreign countries on its Covered List earlier this week, effectively banning the sale of new models in the country without special permission, Taiwanese manufacturer Asus has been quick to respond.

An official statement released by the company yesterday says the following: "Asus has proudly served US customers since 1991, with a long-standing commitment to trusted innovation and strong product security.

"We are confident in the integrity of our supply chain and the security of our networking products. This FCC action has no impact on existing ASUS router users, software updates, and customer support."

With unfortunate timing, Asus also published a major security update for its routers on the very same day. The patch aims to mitigate against CVE-2025-15101, a particularly nasty-sounding vulnerability that scores an 8.5 out of 10 (or High severity) on the official record.

According to its listing, CVE-2025-15101 "potentially allows actions to be performed with the existing privileges of an authenticated user on the affected device, including the ability to execute system commands through unintended mechanisms".

On the one hand, you could argue that the continual discovery and patching of these flaws speaks to Asus' commitment to ongoing security. On the other, the fact that such major vulnerabilities are still cropping up on Asus routers is perhaps a little defeating of the point.

Of course, all router manufacturers experience security issues from time to time, and Asus is far from unique in this regard. Hackers managed to hijack over 16,000 TP-Link routers (and other networking devices) in 2024, creating an Azure-slamming botnet of massive proportions.

Not to mention older Cisco Linksys routers being called out by the FBI for their vulnerabilities, no less. However, Asus does seem to patch its routers more often than most, which again, could be looked at in two different ways.

Certainly, the timing of this particular fix is less than ideal. But hey, it's a good chance for me to remind you once again: Update your networking hardware, folks. It simply ain't worth the risk to ignore.

Asus took the opportunity at CES 2026 earlier this year to unveil two shiny new keyboards: the fast, Hall effect ROG Falchion Ace 75 HE that's focused squarely at gamers; and the one I've had in front of me for the last couple of weeks, the ROG Strix Morph 96 Wireless.

This is a keyboard that's designed to appeal more to the enthusiast end of the mechanical keyboard market, an area we've seen a burgeoning interest in from a range of mainstream manufacturers over recent years. I'd wager Asus has been the one out of all of them to walk the fine line between 'gaming' and 'enthusiast' the best, especially since the Asus ROG Strix Scope II 96 Wireless has sat atop our best gaming keyboards list virtually since its release.

The fact is that the ROG Strix Morph 96 Wireless feels like a strange product when the Strix Scope II 96 Wireless exists. After all, the two share a lot of features. You've got the same 96 percent, or 1800 layout, that squishes up the number pad and moves the nav cluster to provide a full-size keyboard in a reduced footprint, a similar choice of switches, dual wireless connectivity and some very similar physical touches that make this a bit of a head-scratcher to me.

I'll return to this point a little later on, but let's have a gander at what's in front of me. Well, um, it's a keyboard. There you go. I'll take my commission now, please, editor, and be off to the bank to cash my cheque (if I had a physical bank to go to—I guess my phone's banking app will have to do).

Oh, you want more? Oh, alright. I may have to work a little harder to earn a living, I suppose. As I say, we've got the same 96 percent layout as the Strix Scope II 96 Wireless that provides a full-size layout, complete with number pad, arrow keys and nav cluster, in a smaller footprint to save desk space. It's also known as the '1800 layout' after the Cherry G80-1800 keyboard that has been around for at least the last twenty-five years, if not longer by this point, which is where this configuration of keys originated.

It's a functional choice, and one I wish we'd see a little more from manufacturers, being very easy to live with if you're used to a full-size board. You'll just need to remember to find some of the nav keys above the number pad, rather than next to it.

Physical touches include doubleshot keycaps with a pleasant texturing to them that aids grip when typing, although I am quite sad to report they're ABS rather than PBT. This is a marker of a cheaper keyboard, even though PBT has made its way down the price ladder to more wallet-friendly options in recent years, such as the Gamakay x NaughShark NS68. There's potential for the ROG Strix Morph 96 Wireless' keycaps to shine after not too long if you're a heavy user—the Strix Scope II 96 Wireless instead has good old PBT. I am pleased about the dual Windows/macOS legends, though, which is very thoughtful.

The casing here is different to its older brother, too, with a closed shell rather than a skeletal one, which is better because you won't get any ingress of crud, and it keeps the RGB lighting from spilling out. The top case is metal with the underside being plastic, and at 1.125 kg, there's some decent heft here. Unfortunately, I did notice some flex at the corners under some pressure, which leaves a bit of a sour taste. You can also take this keyboard apart with screws in each corner of the top casing, saving the task of turning it over to disassemble. For modders, that's a small but important addition.

On the left, you'll find a multi-function dial that can do everything from controlling volume to changing the RGB lighting brightness level and even scrolling a webpage right out of the box with the function changed by pressing the button further up the left side.

Both sides are home to additional RGB light bars, while the top side houses a USB-C port for wired use, a toggle switch for connectivity and a cubby for the USB-A receiver. The ROG Strix Morph 96 Wireless also comes with some grippy two-stage feet on the underside for added flexibility of typing angle.

Inside, you get a choice of switches, with V2 variants of Asus' linear ROG NX Snow or audibly tactile RX Storm switches; my sample shipped with the former. These are a light and responsive 40 g linear switch that benefits from factory lubrication for no scratchiness under finger. To help their gaming credentials, the ROG NX Snow switches have a slightly higher actuation point and a shorter travel than other switches, for a snappier feel.

For gaming, the ROG NX Snow V2s work well given their lighter force, and have what Asus calls ‘Speed Tap’, which is their version of the SOCD tech found across a lot of manufacturers’ keyboards for unnaturally brisk movement. It is more of a novelty than a feature, given it’s been banned in Counter-Strike 2 online play, for instance.

Beyond this, though, the ROG Strix Morph 96 Wireless lacks the advanced rapid trigger or analog features that are found on similarly priced gaming keyboards with their Hall effect switches, such as the Keychron Q5 HE.

If you don't like the pre-installed switches inside this keyboard, then it is hotswappable, and surprisingly easy to swap out switches and keycaps with the pullers included in the box. To help its enthusiast credentials, this board also comes with a range of foam and such inside for sound dampening, and there isn't any case rattle or ping in my experience. The ROG Strix Morph 96 Wireless is also gasket-mounted, leading to a slightly bouncier feel under finger.

Connectivity is versatile, with both Bluetooth and 2.4 GHz supported for wireless flavour. As is typical, it's essentially plug and play, and I had no issues pairing to Bluetooth on my MacBook Pro and using the 2.4 GHz receiver on my main Windows gaming PC. There is also a USB-C port for wired connectivity and charging. The receiver included with this keyboard is the 'Omni' receiver, which can also handle a 2.4 GHz gaming mouse connection for a potent wireless desktop combo over one receiver, freeing up a valuable USB-A port on your PC.

As for battery life, expect to get around 100 hours of battery life with the RGB lighting set to its default state, and virtually 600 hours if you don't want the RGB on. It is more of a subtle underglow than a full-on Pink Floyd gig, given that the keycaps here have solid legends, so it isn't as much of a sore miss to turn it off as you might expect.

I'm also pleased to report Asus has given the ROG Strix Morph 96 Wireless support for its web-based Gear Link software, which replaces the frankly shoddy Armoury Crate suite. This is only accessible when you have the keyboard plugged in via a cable or the 2.4 GHz receiver, but it provides a much lighter and easier-to-use alternative that gives you a decent amount of customisation.

Here, you can remap keys, fiddle around with the keyboard's wheel, and mess around with the RGB lighting. Granted, you don't get as much functionality as you would with a Hall effect keyboard, but for the reasons you're likely to dive into software, it's adequate.

Returning to the point posed at the beginning of the article, the ROG Strix Morph 96 Wireless feels too similar to the Strix Scope II 96 Wireless in some respects to the point I have considered putting the "They've rebadged it, you fool!" quote from Alan Partridge (British reference, sorry) somewhere in here and therefore gets more of a half-hearted recommendation.

This isn't least as it's possible to get the Strix Scope II 96 Wireless with PBT keycaps, a wristrest and a longer maximum runtime for less than the $140/£135 asking price for this new model.

At this price point, you can also begin to consider the likes of the Ducky Zero 6108, which offers a sublime core experience with a full-size layout, enthusiast-grade touches and proper Cherry MX2A switches for the sub £100/$100 mark. You do have to make do without any software customisation there, though.

It leaves the ROG Strix Morph 96 Wireless in a bit of a funny position. Yes, this is a really lovely keyboard that's cheaper than the Strix Scope II 96 Wireless if we're going on retail price alone with newer switches, a different design and a few other extras, but as I can get other mechanical boards with similar feature sets for less than this, it's a harder sell than you might think.

I'll admit one thing to you right now: I don't like in-ear headphones. I never really have. They just don't feel right in my ears for longer than an hour or so. Maybe it's because I've never been keen on the idea of having a piece of silicon shoved in my ear to listen to things, or the whole form factor, although of course it is all down to taste.

After all, lots of people swear by wireless earbuds for day-to-day productivity loads, and even for gaming. The likes of the SteelSeries Arctis GameBuds and Sony Inzone Buds are two bright sparks in this regard.

Enter the Asus ROG Cetra Open Wireless, a new $230/£185 set of 'open back' gaming buds that could be the answer to my problems. There isn't the need to shove any silicon tips in your ears with these buds, as they simply rest 'on' your ears, rather than in them.

For a quick refresher, open-back headphones refer to ones that don't have closed cups, thereby letting sound in and out for a wider and more expansive presentation at the expense of potentially annoying those around you.

In this instance, the ROG Cetra Open Wireless are designated as 'open' from more of a comfort perspective. It's a strange sensation having them resting on your ears for the first time, especially if you're used to more traditional earbuds and over-ear headphones. In my experience, it took just a few minutes for me to acclimatise to these ROG earbuds, although there was still an odd sensation when I put them back on after taking them off.

Moreover, the inclusion of an ear hook in their construction makes the fit secure and helps position these buds as more of a fitness/running choice, as well as for gaming. Asus also bundles a neck strap in the box to tie the two buds together, so they can't go walkabouts together, and also helps them remain in place if you're running and such. Also helping their fitness credentials is the presence of an IPX5 water resistance rating, meaning these should be alright for water splashes and such from rain or sweat.

One thing I certainly appreciate about these earbuds is that there aren't any finicky touch controls, as Asus has opted for proper buttons instead. Given how small the main frames of these are, the buttons are understandably quite small, requiring a certain degree of precision. Get it right, and you can control media playback and such with relative ease without having to swipe incoherently.

Owing to the larger nature of an open frame, the ROG Cetra Open Wireless' charging case is chunkier than you might be used to—it's more in length than in thickness. The case is still pocketable, just, and you'll need deeper pockets to accommodate. The case has a bit of a plastic feel to it, which is okay, and has some flair with ROG branding if that's your kind of thing.

Open the case up, and you'll find the earbuds sit flat down and magnetically attach for charging, complete with status LEDs so you know juice is going back into them. In the middle sits the USB-C receiver for a low-latency 2.4 GHz connection, complete with a USB-C port in itself for a passthrough so you can still charge your phone while using the buds, or so you don't lose out on the functionality of a front panel USB-C port on your PC, which can be quite hard to come by.

On the point of connectivity, the ROG Cetra Open Wireless connect via both the bundled 2.4 GHz receiver and Bluetooth, and two devices can be connected simultaneously for ease of use. The 2.4 GHz low-latency 'Speednova' connection is handy for gaming powers, while having an additional Bluetooth channel frees these buds up for listening to music or podcasts through your phone when you're travelling, for instance.

Battery life here is generally strong, too, as these Asus earbuds can run for a total of 16 hours from the buds themselves, with the case supplementing this with an additional 48 hours of use. That's 64 hours in total, albeit only if you're happy to turn the mic and RGB lighting off. Leave all of it running, and your runtime will be brought down closer to the SteelSeries Arctis GameBuds' 40 hours, if not slightly less.

A handy feature the ROG Cetra Open Wireless has is wear detection. This is ported over more from 'lifestyle' focused devices than gaming-centric ones, and means the buds will stop whatever you're listening to when it detects you've taken the buds off.

Far too often, I'll take a set of headphones or earbuds off to do something else, only to put them back on and find they've had a party to themselves while I've answered the front door. Now they stop when you do.

It's taken me long enough, but time to get to the important bit: audio quality. Usually, open-back headphones and headsets come with an increasingly wide and expansive presentation—after all, that's the whole point of them—which sets them apart from their closed-back counterparts.

The ROG Cetra Open Wireless largely impressed me in this regard in my usual testing ground of Counter Strike 2, with surprisingly precise directional audio and an excellent soundstage to boot. Combined with this, I got some great overall detail for pinpointing footsteps, distant gunfire and explosions, making these a very competent set of buds for FPS titles.

The big surprise at how these buds sounded, though, was the power of their bass. When not working, I'm usually listening to a good range of rock tracks, be it Marillion, Rush or otherwise. Spinning up Rush's YYZ revealed the punch and power of the bass, providing a great sense of atmosphere and warmth to proceedings. It wasn't to the extent you could feel the subbass as such, but more of a reassuring heft that works away in the background.

I also didn't find that the prominent low end coloured the sound of the ROG Cetra Open Wireless too much, as the mid-range of vocals and footsteps felt clear, while the top-end on the likes of Steely Dan's Do It Again extracted solid detail and sparkle.

I'd probably be inclined to describe these earbuds as a purposeful listen, rather than being a fun or energetic type. The sound signature serves a purpose for their gaming-centric workload, although there is still plenty of room to stretch their legs and have some fun with an array of music.

Even though these are open-back buds, I didn't experience much in the way of seepage from the ROG Cetra Open Wireless. Sure, it was possible to slightly hear the audio from a foot or two away, but that's only at unnaturally loud volumes where you're probably going to do some damage. For the most part, you won't bother people too much using these for a commute or in a busy cafe.

The microphone in these buds is perfectly cromulent for calls and in-game comms, with decent clarity and body overall that I'd wager surpasses a lot of its key rivals.

Then we come to the dilemma of price and appeal. The fact is that the SteelSeries Arctis GameBuds are likely to appeal to more people due to a lower price, a more conventional fit, and as much versatility as these. It's also worth noting you can get some top-class noise-cancelling buds from Sony and Bose for a similar price to these, and more fitness-focused open buds that don't cost as much.

The Asus ROG Cetra Open Wireless are an intriguing addition to the market sector of gaming earbuds, adding the convenience of open-back powers to a form factor it isn't usually associated with. I like the way they fit and the way they sound, although at the price they are, it's an expensive experiment.

The Asus ROG Falchion Ace 75 HE takes the older ROG Falchion Ace HFX and essentially throws it away. That might seem like an overexaggeration, but it seems to be the truth as Asus has promised revamped switches along with a new sensor, new PBT keycaps, a completely different design and more besides. If that isn't starting fresh, I don't know what is.

The brand did this recently with the Azoth 96 HE to great effect, and it seems like this new model has taken some inspiration and elements from it. For instance, we've got the same new ROG HFX V2 switches inside that came on the new Azoth for heightened precision and power against their predecessors.

For reference, the HFX V2 switches have seen some notable upgrades over the older V1 switches, kitted out to previous Asus keyboards with Hall effect powers. The V2s were first seen on the Asus ROG Falcata, with optimisations such as a shorter overall travel (3.5 mm vs 4 mm) that's more noticeable in use than you'd think, plus a larger magnet for more precise customisation. It's now possible to set the actuation or reset point in increments of 0.01 mm, which is not something I've seen before on a range of the best Hall effect keyboards out there.

In conjunction with the optimisations to accuracy and a shorter travel, the ROG HFX V2 switches carry a lighter actuation force of 32 g against 40 g. I appreciate that sounds like a small change, but it's one that makes a big difference. It essentially makes these switches some of the lightest you'll find on a keyboard from a more 'mainstream' brand. Combined with the shorter travel, it makes these some of the speediest out there, and when the rapid trigger powers are taken into account, it can turn the switches into hair triggers.

For the HFX V2 switches, Asus has also moved to a POM stem that's lubricated to mean these linear switches are smooth and slick under finger. To help reduce key wobble, there is also a new block stem—I certainly think these are stable switches under finger, and there isn't much in the way of stem wobble. Asus has also kitted the ROG Falchion Ace 75 HE with five layers of sound dampening inside to nullify any case rattle or ping, and it's a sterling effort here.

The brand also says this keyboard is hotswappable, so it's technically possible to change the switches without the need for any soldering, although it doesn't state which switches work in here besides the HFX V2 and V2X models (the difference between these is that the -X model has an even lighter 30 g force and a -X-style stabiliser assembly for further wobble reduction). It's certainly very rare that you have the option to swap in more traditional three or five-pin MX-style switches in an HE keyboard, after all.

I dived into my usual testing grounds of Counter Strike 2 to give the ROG Falchion Ace 75 HE a fair go, along with the fast Logitech G Pro X2 Superstrike I had spare. The result is that this Asus keyboard is smooth and slick with its inputs, especially with hyper-sensitive rapid trigger enabled for innately brisk actuation and reset points. Movement felt especially fast, as did weapon switching and crouching, owing to the buttons I'd set rapid trigger to in the Gear Link software this keyboard utilises.

Gear Link is Asus' new web-based software that's a major improvement from the app-based Armoury Crate that I've traditionally felt let the ROG products down for far too long, feeling slow and clunky. Being web-based means Gear Link feels lighter and more responsive for its impressive array of features, including setting actuation and reset points, plus rapid trigger powers, and more advanced mapping such as dynamic keystroke, mod tap and all that jazz if you want to fiddle until your heart's content.

There is also SOCD tech for unnaturally quick strafing, known here as Speed Tap, although this is more for marking bumf than for real-world use. Use it online, and you'll probably see the almighty ban hammer wielded. You can also fiddle with RGB lighting, plus with the functionality of this keyboard's touch surface on the top side and the roller wheel near the end of the right-hand edge for things such as key travel, RGB lighting brightness and media volume.

The ROG Falchion Ace 75 HE is entirely wired, meaning a convenient and latency-free connection, plus there's no need to worry about charging it up. You also get the benefit of 8000 Hz polling powers, too, for more frequent reporting of inputs than other, more standard gaming keyboards. It might not be something that mere mortals can benefit from, but for the pros that this board is aimed at, where, to them, every fraction of a second counts, it's a worthwhile addition.

The design of this Asus keyboard has changed a little against the older Falchion models, with this variant moving up to a 75% layout for the added functionality of a function row and single-column nav cluster. The Falchion's trademark touch bar across the top makes a return with this model, as well as a roller in the top right corner. Both can be remapped in software for a variety of functions and give this keyboard a clever means of control besides the actual keyboard part.

The roller is a particularly ingenious addition, as by default, it's used for fine-tuning actuation point and the sensitivity of the keyboard's rapid trigger without going into software. Plus, there's a toggle switch on the right side that can quickly enable and disable rapid trigger at the flick of a switch.

The fit and finish here is suitably excellent, with Asus' customary blend of a plastic underside and metal top plate to the chassis for a sturdy feel in-hand. It is possible to get fully machined aluminium cases at this price point, but this is a good mix of materials. In addition, this keyboard has doubleshot PBT keycaps for extra durability and a quality touch.

Then, after all this, we come to the price. The ROG Falchion Ace 75 HE is going to leave a £220/$220-shaped hole in your wallet, making it a real premium product. At that price, we're looking at rivals including the larger layout Keychron Q5 HE that has a fully machined aluminium case at the expense of lacking slightly more advanced switches and the low-profile NuPhy Air60 HE that trades full-height switches for low-profile ones and retains a lot of the same powers at a lower price tag.

There is also another thorn in the ROG Falchion Ace 75 HE's armour, with one of our favourite HE keyboards, the wireless Keychron K2 HE, which trades wired power for wireless convenience and adds some pleasant wood to the chassis if you get the right version. It too has powerful HE switches, convenient software, and will keep more pounds/dollars/whichever currency in your pocket while doing so.

The bottom line is simple: the Asus ROG Falchion Ace 75 HE is an excellent HE gaming keyboard, but is beaten by more affordable competition that doesn't sacrifice against it. Bear in mind the usual Asus ROG tax before picking this one up.

The last time I gamed on a 4K 27-inch display, what I really wanted was even more pixels. Over eight million obviously just ain't enough.

Actually, no I absolutely didn't. But the new Asus ROG Strix XG27JCG is giving them to me anyway, all 15 million of tiny RGB suckers thanks to a 5K native resolution of 5,120 by 2,880 pixels from its 27-inch IPS panel. The result is frankly epic pixel density, plus a whole raft of questions about what, exactly, the point of this kind of display is for gaming.

But before you dismiss the idea of a 5K gaming monitor out of hand on account of low refresh rates, know that the ROG Strix XG27JCG will do fully 180 Hz at the 5K native. Okay, that's pretty pedestrian by the standards of really quick gaming monitors, which clock in upwards of 500 Hz these days. But still, 180 Hz at 5K. Bananas, isn't it?

Actually, this monitor's high-refresh chops extend beyond that thanks to a 1440p dual-mode functionality that supports 330 Hz. The idea here is the same as 4K dual-mode monitors that offer an alternative 1080p "native" at a higher refresh, like Asus' own ROG Strix XG27UCG. But since this is a 5K rather than 4K panel, if you double up the pixels in both horizontal and the vertical—effectively using a square block of four pixels from the 5K grid to form a single pixel—you end up with an alternative 1440p "native" mode of 2,560 by 1440.

So, yeah, this monitor is basically four 1440p panels with a total of 14,745,600 pixels. That's how ridiculously high the resolution is with 5K. But with that dual-mode functionality, in theory you can have all the benefits of 5K—which we'll come to—plus a more realistic 1440p mode that promises to make this monitor usable with vaguely affordable GPUs and also a decent option for esports. After all, running at the full native 5K native in the latest games is a very big ask in terms of raw PC performance, not to mention largely pointless for online shooters where frame rate rather than image detail is king.

The 5K and dual-mode functionality aside, however, this is a fairly mainstream monitor by most other metrics. As I mentioned, it's an IPS panel rather than OLED and while it is HDR 600 certified and has local dimming, it's limited to 10 edge-lit zones and so HDR support is very basic indeed. That said, the HDR 600 certification at least means that the panel can hit 600 nits, which is plenty bright enough. The exception is the 0.3 ms GtG claimed pixel response time. That's quick for an IPS monitor.

Rounding out the feature set is good connectivity thanks to DisplayPort, HDMI and USB-C, the latter with basic 15 W power delivery which is probably good enough for sharing this screen between a gaming desktop and a thin-and-light work laptop, or similar. Just note that the full 180 Hz at 5K isn't available over USB-C, you'll need to use one of the other interfaces for that.

Anyway back to that 5K native res and the nearly 15 million pixels. What, you might ask, is the point? Surely 4K on a 27-inch panel is plenty, especially for gaming. And you'd definitely have a point. But, for the record, the increased pixel density on offer here, which comes in at 215 pixels per inch versus the 163 pixels per inch of a regular old 287-inch 4K monitor, is genuinely quite stunning at times.

Put it this way, you know those crisp, clean fonts that you're used to from your smartphone? Well, you get something approaching that on your desktop PC thanks to this monitor. Actually, it's not quite that good in Windows simply because that operating system's font rendering is a bit clunky. But plug this monitor into an Apple Mac and then you really do get that smartphone experience when it comes to text. A niche attraction, to be sure, but there you have it.

Whatever OS you use, there's huge image detail on offer. Indeed, you can view 4K content at full resolution in a window, leaving space for toolbars and controls. That's appealing for, say, 4K video editing. Actually, this screen reveals flaws in 4K video streams that you never noticed before; there's that much detail on offer.

As for the broader utility of 5K, well, in theory it gives huge desktop space. But in practice, running at 100% scaling would make everything ludicrously tiny. So your desktop elbow room will depend on your favoured scaling. In other words, there's no real advantage over a 4K monitor in that regard.

All of which leaves us with gaming. And that's actually quite a complicated proposition on this panel. The thing is, 5K isn't as preposterous a target resolution as it would have been a few years ago. Partly, that's thanks to more powerful GPUs, but mostly it's thanks to upscaling technologies such as Nvidia's DLSS.

The incredible thing is that, thanks to DLSS, you can get pretty playable results at 5K using something like an Nvidia RTX 4070 or RTX 5070. Those aren't budget GPUs, for sure, but they're also not $2,000 monsters. With Nvidia latest DLSS Transformer model upscaling, Performance mode looks remarkably good at 5K and gives you a base frame rate in something like Cyberpunk 2077 with most of the visual bells and whistles cranked up roughly in the low 30s with, say, an RTX 5070.

That's tolerable in latency terms, so you can then add some frame gen and end up with genuinely playable frame rates at 5K. Pretty impressive, right? Step up to, for instance, an RTX 5080 and that 5K native only gets more usable. Of course, you also have the 1440p alternative mode. I had high hopes for that given that 1440p on a 27-inch panel is still reasonably high resolution in gaming terms. But the results are a teensy bit disappointing.

As with basically every other dual-mode monitor I've tried, there's no mistaking it for an actual 1440p panel running at native. The image is somehow both softer and subjectively a bit more pixelated than that. Funnily enough, I think the 1440p alt mode actually looks better running some upscaling than it does running pure "native". Consequently, I think this panel probably doesn't use straight-up pixel doubling and instead has some kind of interpolation algorithm. Again, that seems to be the case for most if not all dual-mode monitors.

Of course, if you're using the 1440p mode for its higher refresh for esports as opposed to reduced GPU load for really demanding games, the finer detail of image quality probably doesn't matter. The only thing I'd say in that context is that if online shooters are your priority, the problem is that you can get a higher refresh and faster responding 1440p OLED at this price point, and that's got to be a better fit.

In the end, it's hard to draw a really clear conclusion about the whole 5K-with-dual-mode thing. On the one hand, the 5K native resolution is actually more usable than you might think thanks to modern upscaling technologies. On the other, the 1440p alternative mode isn't quite as crisp and sharp as you'd ideally want. Still, you can't argue that this monitor doesn't give you plenty of options.

It's also a really nice monitor is most other regards. The calibration is basically bob-on in every mode, be that sRGB and DCI-P3 SDR modes or with HDR enabled. As I said, this monitor doesn't have proper local dimming. So, the actual HDR capability is essentially negligible. But at least SDR content calibration in HDR mode is also really nicely judged. So, there's no real downside to enabling HDR.

So, this is a really nice IPS monitor with fantastic colours, plus that epic pixel density for razor-sharp image detail. The problem is that at this price point, there are also plenty of OLED options, and in that context the limited contrast is all too obvious. As with basically any IPS monitor, there's a lot of backlight bleed through the panel, leading to poor black levels. That's LCD tech for you.

Personally, I think that's less of an issue on a relatively compact monitor like this. When you have a really massive ultrawide monitor, backlight bleed and indeed the variability in how visible that can be across different parts of the panel, again thanks to the inherent viewing angle limitations of LCD tech, can be quite distracting. On a 27 incher, it's not quite as annoying.

And yet this is a very expensive display that ends up competing, as I said, with plenty of OLED options on price. Strictly speaking, none of those OLEDs can match this LCD when it comes to pixel density. But as I write these words, you can grab a 27-inch 4K OLED for about $200 less or a 144 Hz 27-inch 4K IPS panel for less than one-third the money. Either way, the alternatives end up making this 5K monitor, as impressive as it is in many ways, seem like a problematic value proposition.

The memory crisis appears to be hitting almost everyone, from PC gamers to PC suppliers, but Asus doesn't seem too worried. Addressing concerns in its latest financial earnings report, Asus reckons it is uniquely positioned to survive it.

Opening the topic, Asus CFO Nick Wu says, "I would like to briefly add that as investors are currently focusing on the rise in memory prices in 2026, and of course, that may suppress the demand and consequently the profitability in the 2026 PC market."

However, he argues that there are three major elements that "provide a very good opportunity to examine Asus's brand value and operational resilience"

The first is that Asus is a high-end brand with "high-end brand value and high-end product mix." Wu doesn't go into detail about how, specifically, Asus being a high-end brand will help with those rising memory costs, but one could assume that, being a well-known name, some consumers might opt to pay more for its desktops and laptops. The product mix is a notable nod too, as margins elsewhere will stay consistent with pre-memory crisis ones.

The second point Asus makes is that partnerships held with operational partners "is a critical component to our overall operational resilience, and it does pay dividends at such times as right now."

Asus, being a trusted brand, will have contracts and relationships that could potentially get it hardware before less established companies. This can actually make the memory crisis a bit of an opportunity for some brands, as they can withstand it through tactical partnerships, in ways others cannot.

The last element that Asus points to is the fact that it also manufactures graphics cards and motherboards. The former category has been shooting up in price, and the latter one will dip if gaming PC demand goes down. Diversifying its range of products is naturally good for market consistency, and like gaming PCs, Asus can pass the extra price of manufacturing products onto the consumer.

Wu says, "Overall, we are very confident in achieving a stable operation in 2026 that will be able to outperform the industry."

One significant way that Asus plans on increasing revenue going forward is by further committing to, you guessed it, AI. Both the cause of the memory crisis and part of Asus' route around it, co-CEO S.Y. Hsu says, "Asus is fully riding the AI wave to take off."

Well, it seems like companies giving up on AI and letting memory prices fall back to normal isn't on the cards in the near future.

Even if you've sufficiently managed your expectations and aren't going for the flashiest bit of kit, hardware is still expensive. Full desktop rigs are bad enough, but even mid-range gaming laptops can come at an eye-watering premium—and don't even get me started on just how much an Apple alternative would set you back by.

MacBooks are expensive and their gaming performance still isn't what you'd want, right? Well, for one thing Apple is a better PC gaming platform than people give it credit for, and now the just announced MacBook Neo is presenting an affordable inroad to its very shiny walled garden, for only $599.

Powered by the A18 Pro mobile chip and paired with 8 GB of RAM, the MacBook Neo is still far from the most powerful laptop out there. The A18 Pro chip debuted in the iPhone 16 Pro in 2024, so it's hard not to think of this as an oversized phone with a keyboard and a pretty swish 13-inch Liquid Retina display.

The very competitive phone-like pricing does little to argue against that notion either. For $599, I wouldn't be surprised if this became the student device of choice over, say, an iPad or a similar tablet device—the fun colours certainly help.

That said, someone on a student budget could potentially still net something with a little more grunt. For instance, the Samsung Galaxy Book4 Edge enjoys the power of the Snapdragon X chip and 16 GB of RAM for close to $700—or you could scoop up the same chip and generous helping of RAM in the Asus Vivobook 14 for only $500. If you're deadset on Apple, there's also the very reasonably priced Mac Mini (the M4 Pro chip inside makes it a decent gaming option).

Apple are pitching the MacBook Neo as 'a powerful platform for AI with Apple Intelligence built right in,' but only 8 GB of RAM is a pretty meagre offering for anyone really wanting to throw down with larger generative models.

It's fair to say Apple has more lightweight AI-use in mind, claiming the Neo is "3x faster when running on-device AI workloads like applying advanced effects to photos, compared to the bestselling PC with the latest shipping Intel Core Ultra 5."

Unfortunately, whatever battery gains Apple could have made with the mobile chip are somewhat squandered by the 36.5 Whr battery on board. Apple claims you'll still get about 16 hours out of it—but I doubt that's 16 hours of playing Kpop Demon Hunters on repeat and at full volume. 'So sweet, so easy on the eye, but hideous on the inside—' What? It's still a banger.

As much as I like the sound of bedtime gaming, I haven't yet committed to a handheld gaming PC of my own. Part of that is down to pricing, but a not insignificant part of it is also to do with the question of long-term hardware support.

Lenovo, for one, is saying it's in the handheld arena for the long haul. Specifically, the company said it planned to support its original AMD Ryzen Z1 Extreme-based handheld, the Lenovo Legion Go, until at least 2029.

"Support for the Lenovo Legion Go (8.8”, 1) has not been discontinued," The company wrote in an official statement issued to PCWorld, "Lenovo is actively continuing to support the Legion Go (8.8”, 1) with necessary driver and BIOS updates and will continue to do so through October 2029. Lenovo is working in concert with AMD on driver update cadence, and new updates will be released once they have passed Lenovo’s rigorous review protocols."

Concerns around long-term support for the Legion Go first arose after a machine translation of a Lenovo Korea customer service response suggested that there were no plans for the original handheld to receive any further driver updates. At the time, the Lenovo Legion Go hadn't received a GPU driver update since last September.

Last week, our Nick investigated another similarly AMD Ryzen Z1 Extreme-based handheld, the Asus ROG Ally. The handheld's attempt to run Resident Evil Requiem left much to be desired, especially compared to the older Steam Deck.

A driver update arrived for the ROG Ally the same week as Nick's story, though it did not substantially improve the handheld's in-game performance. The entire episode did not seem to disprove the rumour that AMD had abandoned driver support for its Ryzen Z1 chips.

Lenovo's latest statement is highly specific and declarative, instead suggesting a continuing support commitment to the Ryzen Z1 Extreme chip from AMD. That's just as well, as the original Lenovo Legion Go is barely two years old and can still go for a premium price even second-hand. However, it's interesting to note that to date we've not heard anything quite as confident from Asus ROG about support for the Ally.

Last week, both Asus and Acer were blocked from directly selling gaming PCs in Germany due to a lawsuit from phone manufacturer Nokia. This not only leaves potential customers with fewer choices when it comes to their next rig, but now that the websites are down, they are left without updates too. Or at least without some extra effort.

However, Acer has informed me: "The site should be up again shortly but in the meantime German customers can contact Acer support via the website, or alternatively also download updates here."

As pointed out by ComputerBase, trying to access the German websites for Acer and Asus will greet you with an error message. It seems like this is reliant on your geolocation, so using a VPN will get you around the errors, but it's a sign of how things have progressed since Nokia's case.

Interestingly, ComputerBase also notes that you can access the Asus site in Germany by going through the Chinese version of the site. Users of Acer gear can get around restrictions by using the Swiss Acer support page, then finding whatever update they need from there.

As a chronic 'I swear I'll update it later' PC user, any steps put in between me and my updates will only make me more likely to procrastinate, so I feel for the German public here. You can potentially still buy gaming PCs in Germany, just not directly, so updating will still be a necessary thing for users.

The lawsuit from Nokia alleges that Asus and Acer's PCs (alongside Hisense's smart TV) infringe three of its patents. They are related to the H.265 (HEVC) video code standard. Though Hisense settled the case in January, Asus and Acer are still fighting it. The blocking of direct sales is an injunction from the Munich court.

Notably, the lawsuit does not include monitors, peripherals or graphics from either brand in the region, so the sites going offline is particularly strange.

Users who have bought items not hit in the lawsuit from either brand are also seeing a negative effect here. With Acer set to send a new website live, here's hoping Asus follows suit.

I have also reached out to Asus for comment and will update this story if I hear back.

The ROG Xbox Ally X was already expensive when it launched late in 2025, but its Australian RRP of AU$1,599 has just increased to AU$1,799. That AU$200 increase is pretty significant when you consider that other consoles, like the PlayStation 5 and Xbox Series X, usually get price hikes in increments of AU$30 to AU$50.

In other words: If you wanted a ROG Xbox Ally X and didn't buy one yesterday, you're probably kicking yourself. I've reached out to Asus for more information on the price increase, and will update when I hear back. The price rise is reflected by all retailers that sell the unit directly in Australia, including JB Hi-Fi and ASUS direct.

It's pretty easy to put the pieces together ourselves though. The RAM and storage crisis is in full swing and few gaming hardware makers are unaffected by it. It's affecting Steam Deck OLED supply in some regions, and it's reportedly prompting Nintendo to reconsider its price for the Nintendo Switch 2. Sony is also reportedly looking to delay the PS6, while Valve has confirmed its Steam Machine and Steam Frame hardware has been pushed back due to the shortages.

The Ally X is one of two Xbox-branded handhelds released last year. It's the more powerful device: the lesser ROG Xbox Ally hasn't received a price increase from its AU$999 RRP, and in fact, can usually be had for at least AU$100 cheaper (such as what Amazon is offering at the moment). Dave James reviewed the Ally X last year and liked it, but noted even then that it's a pricey beast.

According to VGC, the ROG Xbox Ally X has also copped a price increase in Japan of the equivalent of AU$200.

The RAM and storage crisis can be mostly blamed on the AI boom, which is sucking up all available resources so that blue-tick X users can generate cringe-inducing approximations of Star Wars and declare the film industry is over. The CEO of Phison thinks many consumer electronics manufacturers "will go bankrupt or exit product lines" by the end of the year thanks to the crisis.

German PC gamers might struggle to pick up Acer or Asus gaming laptops and desktops for the foreseeable future, as a court ruling out of Munich has put a pause on direct sales. This is the latest development in a patent suit filed by Nokia against the two companies, as well as Chinese electronics company Hisense last year.

The lawsuit alleges that Acer and Asus’ PCs plus Hisense’s smart TVs infringe upon three of Nokia's patents relating to the H.265 (HEVC) video codec standard (via Juve Patent). HiSense sought to settle in January and, following a license agreement, Nokia has since withdrawn its lawsuit against the company. Acer and Asus are still fighting it out in the courts.

This latest Munich court ruling is an injunction against Acer and Asus, but not retailers. That means German gamers won't be able to buy directly from either company, but should still be able to find Acer and Asus machines elsewhere. Acer assured German-language outlet PC-Welt that its other products, from monitors to even e-scooters, were unaffected by this ruling.

Acer did not want to comment on ongoing legal proceedings, but it's worth noting the company has also filed its own separate suit against Nokia at the Munich Regional Court, though this instead regards wireless communication tech.

Legal proceedings between the three companies are not confined to the German court system either. Back in December, the UK High Court awarded Acer and Asus interim licenses, though Nokia has also filed a separate suit in the US alleging infringement of its Standard Essential Patents (SEPs). Basically, this is far from the end of it.

Aside from patent proceedings, Nokia has most recently earned headlines with 4G projects that shoot for the moon—literally.

On the other side of the suit, Asus has most recently been in the news for announcing an 'immediate internal review' of its 800-series motherboards following 'recent reports concerning Ryzen 7 9800X3D CPUs.' To elaborate on the company's vaguely worded statement on the issue, some PC gamers allege that popping a Ryzen 7 9800X3D CPU into an AM5 socket motherboard has caused a whole host of issues, up to and including the death of the CPU itself.

Perhaps hitting the pause button on planned Asus purchases is not the worst thing a consumer could do right now.

The Asus GeForce RTX 5080 Noctua Edition is claimed to be the "quietest graphics card in its class" and I don't doubt it's in the running. Though upon first removing it from its box, I was sure it had earned itself at least one accolade: it's easily the largest graphics card I've ever tested. Just a year or two ago, I would have laughed at sheer heft of the RTX 4080 Founders Edition—it looks practically pint-sized next to the Noctua Edition.

This colossus is the product of a partnership between Asus and Noctua. The cooling company is providing three of its NF-A12x25 120 mm fans for the shroud here, including a splash of brown that we don't see on a GPU too often. I think it looks great, with an almost steampunk appearance on account of its massive heft and unusual colourway. It even sparkles a little in the light, which is a nice touch.

Though its powerful look is a limit on what you can pair with it. It's very easy to clash given it's sporting Noctua's telltale brown, unless you pair it with more Noctua brown.

So that's what I did. I paired the RTX 5080 Noctua Edition with the only kit I could think to: the Noctua NH-D15 G2 cooler and a couple NF-A14x25 G2 fans. All loaded into the Havn BF 360. A gorgeous combination, if I dare say myself.

But it's also extremely easy to build an ugly PC with this card. I moved it over to our MSI/Cyberpower test system for benchmarking, which is all-white and bathed in RGB lights, and the RTX 5080 looks absolutely horrid. I had to avert my eyes for the remainder of the tests.

Weighing in at 2.607 kg—enough that I weighed it on my human scales, not my kitchen scales—this card features a huge heatsink riddled with heat pipes. Four of these heat pipes run the length of the card, which, at 385 mm, is only able to fit inside very spacious cases. The rest culminate near the vapour chamber that sits atop the GPU.

One of the less spoken about but important parts of a graphics card's construction is the thermal interface material (TIM). This card offers a phase-change thermal pad to keep an even contact between the heatsink and GPU at all times, and ideally, over a long period of time. The card has a lapped contact plate atop of the GB203 GPU, a technique that Asus calls MaxContact. It reckons this alone knocks up to 2 °C off.

For I/O, this card includes two HDMI 2.1 ports (one above the other) alongside three DisplayPort 2.1 ports, which run the remaining length of the two-slot bracket. It's absolutely a four-slot card, however, measuring 388 x 80 x 138 mm by my own measurements

The graphics card offers two BIOS modes: Performance and Quiet. These are switched via a tiny switch near the 16-pin connector. The Quiet mode keeps the clock speed at 2,700 MHz. The Performance mode sets it to 2,730 MHz.

Performance - Asus GeForce RTX 5080 Noctua OC Edition

3DMark Port Royal3DMark Time Spy ExtremeBlack Myth Wukong (1080p High)Black Myth Wukong (1440p High + Upscaling (quality))Black Myth Wukong (1440p High)Black Myth Wukong (4K High)Blender 4.3.0 (Junkyard)Cyberpunk 2077 (1080p RT Ultra)Cyberpunk 2077 (1440p RT Ultra + upscaling (quality))Cyberpunk 2077 (1440p RT Ultra)Cyberpunk 2077 (4K RT Ultra)F1 24 (1080p Ultra high)F1 24 (1440p Ultra High)F1 24 (1440p Ultra high +upscsaling (Quality), 2x FG)F1 24 (4K Ultra High)Homeworld 3 (1080p Epic)Homeworld 3 (1440p Epic + upscaling (quality))Homeworld 3 (1440p Epic)Homeworld 3 (4K Epic)Metro Exodus Enhanced Edition (1080p Ultra)Metro Exodus Enhanced Edition (1440p Ultra)Metro Exodus Enhanced Edition (4K Ultra)Procyon generative image creationThe Talos Principle 2 (1080p Ultra)The Talos Principle 2 (1440p Ultra +upscaling (quality), 2x FG)The Talos Principle 2 (1440p Ultra)The Talos Principle 2 (4K Ultra)

RTX 5080 Asus Noctua Edition (performance mode)

22443

RTX 5080 Asus Noctua Edition (quiet mode)

22940

RTX 5080 MSI Ventus 3X OC White

21669

07,50015,00022,50030,000

Overall index score

3DMark Port Royal Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	22443
RTX 5080 Asus Noctua Edition (quiet mode)	22940
RTX 5080 MSI Ventus 3X OC White	21669

GPU index score

Overall index score

RTX 5080 Asus Noctua Edition (performance mode)

16311

13779

RTX 5080 Asus Noctua Edition (quiet mode)

16353

13738

RTX 5080 MSI Ventus 3X OC White

15675

13452

05,00010,00015,00020,000

3DMark Time Spy Extreme Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	16311 GPU index score, 13779 Overall index score
RTX 5080 Asus Noctua Edition (quiet mode)	16353 GPU index score, 13738 Overall index score
RTX 5080 MSI Ventus 3X OC White	15675 GPU index score, 13452 Overall index score

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

121

93

RTX 5080 Asus Noctua Edition (quiet mode)

122

104

RTX 5080 MSI Ventus 3X OC White

123

106

037.575112.5150

Black Myth Wukong (1080p High) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	121 Avg FPS, 93 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	122 Avg FPS, 104 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	123 Avg FPS, 106 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

130

111

RTX 5080 Asus Noctua Edition (quiet mode)

131

114

RTX 5080 MSI Ventus 3X OC White

133

114

037.575112.5150

Black Myth Wukong (1440p High + Upscaling (quality)) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	130 Avg FPS, 111 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	131 Avg FPS, 114 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	133 Avg FPS, 114 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

93

81

RTX 5080 Asus Noctua Edition (quiet mode)

93

82

RTX 5080 MSI Ventus 3X OC White

92

81

0255075100

Black Myth Wukong (1440p High) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	93 Avg FPS, 81 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	93 Avg FPS, 82 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	92 Avg FPS, 81 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

56

49

RTX 5080 Asus Noctua Edition (quiet mode)

56

49

RTX 5080 MSI Ventus 3X OC White

54

48

015304560

Black Myth Wukong (4K High) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	56 Avg FPS, 49 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	56 Avg FPS, 49 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	54 Avg FPS, 48 1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

2161.54

RTX 5080 Asus Noctua Edition (quiet mode)

2139.47

RTX 5080 MSI Ventus 3X OC White

2284.56

07501,5002,2503,000

Samples per min

Blender 4.3.0 (Junkyard) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	2161.54
RTX 5080 Asus Noctua Edition (quiet mode)	2139.47
RTX 5080 MSI Ventus 3X OC White	2284.56

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

104

75

RTX 5080 Asus Noctua Edition (quiet mode)

103

81

RTX 5080 MSI Ventus 3X OC White

103

76

037.575112.5150

Cyberpunk 2077 (1080p RT Ultra) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	104 Avg FPS, 75 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	103 Avg FPS, 81 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	103 Avg FPS, 76 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

110

79

RTX 5080 Asus Noctua Edition (quiet mode)

110

86

RTX 5080 MSI Ventus 3X OC White

182

126

050100150200

Cyberpunk 2077 (1440p RT Ultra + upscaling (quality)) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	110 Avg FPS, 79 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	110 Avg FPS, 86 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	182 Avg FPS, 126 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

68

55

RTX 5080 Asus Noctua Edition (quiet mode)

66

55

RTX 5080 MSI Ventus 3X OC White

70

56

020406080

Cyberpunk 2077 (1440p RT Ultra) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	68 Avg FPS, 55 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	66 Avg FPS, 55 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	70 Avg FPS, 56 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

35

28

RTX 5080 Asus Noctua Edition (quiet mode)

34

28

RTX 5080 MSI Ventus 3X OC White

34

28

010203040

Cyberpunk 2077 (4K RT Ultra) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	35 Avg FPS, 28 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	34 Avg FPS, 28 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	34 Avg FPS, 28 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

175

126

RTX 5080 Asus Noctua Edition (quiet mode)

170

128

RTX 5080 MSI Ventus 3X OC White

180

132

050100150200

F1 24 (1080p Ultra high) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	175 Avg FPS, 126 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	170 Avg FPS, 128 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	180 Avg FPS, 132 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

120

95

RTX 5080 Asus Noctua Edition (quiet mode)

118

95

RTX 5080 MSI Ventus 3X OC White

123

98

037.575112.5150

F1 24 (1440p Ultra High) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	120 Avg FPS, 95 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	118 Avg FPS, 95 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	123 Avg FPS, 98 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

178

129

RTX 5080 Asus Noctua Edition (quiet mode)

177

130

RTX 5080 MSI Ventus 3X OC White

266

156

075150225300

F1 24 (1440p Ultra high +upscsaling (Quality), 2x FG) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	178 Avg FPS, 129 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	177 Avg FPS, 130 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	266 Avg FPS, 156 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

62

49

RTX 5080 Asus Noctua Edition (quiet mode)

62

50

RTX 5080 MSI Ventus 3X OC White

64

50

020406080

F1 24 (4K Ultra High) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	62 Avg FPS, 49 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	62 Avg FPS, 50 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	64 Avg FPS, 50 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

149

68

RTX 5080 Asus Noctua Edition (quiet mode)

150

70

RTX 5080 MSI Ventus 3X OC White

145

67

037.575112.5150

Homeworld 3 (1080p Epic) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	149 Avg FPS, 68 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	150 Avg FPS, 70 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	145 Avg FPS, 67 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

144

69

RTX 5080 Asus Noctua Edition (quiet mode)

149

69

RTX 5080 MSI Ventus 3X OC White

141

66

037.575112.5150

Homeworld 3 (1440p Epic + upscaling (quality)) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	144 Avg FPS, 69 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	149 Avg FPS, 69 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	141 Avg FPS, 66 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

132

66

RTX 5080 Asus Noctua Edition (quiet mode)

136

68

RTX 5080 MSI Ventus 3X OC White

129

67

037.575112.5150

Homeworld 3 (1440p Epic) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	132 Avg FPS, 66 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	136 Avg FPS, 68 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	129 Avg FPS, 67 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

103

65

RTX 5080 Asus Noctua Edition (quiet mode)

102

66

RTX 5080 MSI Ventus 3X OC White

102

65

037.575112.5150

Homeworld 3 (4K Epic) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	103 Avg FPS, 65 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	102 Avg FPS, 66 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	102 Avg FPS, 65 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

188

54

RTX 5080 Asus Noctua Edition (quiet mode)

190

55

RTX 5080 MSI Ventus 3X OC White

184

55

050100150200

Metro Exodus Enhanced Edition (1080p Ultra) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	188 Avg FPS, 54 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	190 Avg FPS, 55 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	184 Avg FPS, 55 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

148

49

RTX 5080 Asus Noctua Edition (quiet mode)

147

49

RTX 5080 MSI Ventus 3X OC White

140

48

037.575112.5150

Metro Exodus Enhanced Edition (1440p Ultra) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	148 Avg FPS, 49 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	147 Avg FPS, 49 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	140 Avg FPS, 48 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

91

41

RTX 5080 Asus Noctua Edition (quiet mode)

91

41

RTX 5080 MSI Ventus 3X OC White

85

40

0255075100

Metro Exodus Enhanced Edition (4K Ultra) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	91 Avg FPS, 41 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	91 Avg FPS, 41 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	85 Avg FPS, 40 1% Low FPS

Index score

Duration (sec)

RTX 5080 Asus Noctua Edition (performance mode)

5196

19.242

RTX 5080 Asus Noctua Edition (quiet mode)

5192

19.257

RTX 5080 MSI Ventus 3X OC White

4770

4770

01,5003,0004,5006,000

Procyon generative image creation Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	5196 Index score, 19.242 Duration (sec)
RTX 5080 Asus Noctua Edition (quiet mode)	5192 Index score, 19.257 Duration (sec)
RTX 5080 MSI Ventus 3X OC White	4770 Index score, 4770 Duration (sec)

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

122

85

RTX 5080 Asus Noctua Edition (quiet mode)

123

85

RTX 5080 MSI Ventus 3X OC White

123

63

037.575112.5150

The Talos Principle 2 (1080p Ultra) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	122 Avg FPS, 85 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	123 Avg FPS, 85 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	123 Avg FPS, 63 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

125

84

RTX 5080 Asus Noctua Edition (quiet mode)

125

84

RTX 5080 MSI Ventus 3X OC White

195

101

050100150200

The Talos Principle 2 (1440p Ultra +upscaling (quality), 2x FG) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	125 Avg FPS, 84 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	125 Avg FPS, 84 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	195 Avg FPS, 101 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

91

64

RTX 5080 Asus Noctua Edition (quiet mode)

91

64

RTX 5080 MSI Ventus 3X OC White

90

61

0255075100

The Talos Principle 2 (1440p Ultra) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	91 Avg FPS, 64 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	91 Avg FPS, 64 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	90 Avg FPS, 61 1% Low FPS

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (performance mode)

52

37

RTX 5080 Asus Noctua Edition (quiet mode)

52

37

RTX 5080 MSI Ventus 3X OC White

52

36

015304560

The Talos Principle 2 (4K Ultra) Data

Product	Value
RTX 5080 Asus Noctua Edition (performance mode)	52 Avg FPS, 37 1% Low FPS
RTX 5080 Asus Noctua Edition (quiet mode)	52 Avg FPS, 37 1% Low FPS
RTX 5080 MSI Ventus 3X OC White	52 Avg FPS, 36 1% Low FPS

PC Gamer test platform
Supplied by Cyberpower | MSI

CPU: AMD Ryzen 7 9800X3D | Motherboard: MSI MPG X870E Edge Ti WiFi | RAM: Kingston Fury Beast RGB 32 GB (2 x 16 GB) @ 6,000 MT/s | Cooler: MAG CoreLiquid i360 White | SSD: Spatium M480 Pro 2 TB | PSU: MPG A1000GS PCIe 5 | Case: MAG Pano 100R White

Switching between the Quiet and Performance modes doesn't result in any major changes for performance or noise levels. A frame or two in Cyberpunk 2077 at most. That checks out, not only because of the small increase in clock speed between the two, but how much further the RTX 5080 can be pushed with an actual overclock. You can slap a couple hundred MHz more without even thinking about it—let alone with a beefy card such as this.

In fact, I slapped on a 400 MHz core offset and a 1,000 MHz memory offset onto the Noctua without thinking about it. I added an extra 10% power, too—the Noctua Edition offers up to 25%. In Metro Exodus Enhanced Edition, that earned me eight more frames on average right there. I barely lifted a finger.

Overclocking - Asus GeForce RTX 5080 Noctua OC Edition

Metro Exodus Enhanced Edition (4K Ultra)Performance per wattThermalsPowerGPU frequency

Avg FPS

1% Low FPS

RTX 5080 Asus Noctua Edition (+400 core, +1000 mem, +10% PL)

99

44

RTX 5080 Asus Noctua Edition (performance mode)

91

41

0255075100

Metro Exodus Enhanced Edition (4K Ultra) Data

Product	Value
RTX 5080 Asus Noctua Edition (+400 core, +1000 mem, +10% PL)	99 Avg FPS, 44 1% Low FPS
RTX 5080 Asus Noctua Edition (performance mode)	91 Avg FPS, 41 1% Low FPS

4K frames per J

1080p frame per J

RTX 5080 Asus Noctua Edition (+400 core, +1000 mem, +10% PL)

0.27

0

RTX 5080 Asus Noctua Edition (performance mode)

0.27

0.71

00.20.40.60.8

Performance per watt Data

Product	Value
RTX 5080 Asus Noctua Edition (+400 core, +1000 mem, +10% PL)	0.27 4K frames per J, 0 1080p frame per J
RTX 5080 Asus Noctua Edition (performance mode)	0.27 4K frames per J, 0.71 1080p frame per J

Peak temperature (°C)

Average temperature (°C)

RTX 5080 Asus Noctua Edition (+400 core, +1000 mem, +10% PL)

62

58

RTX 5080 Asus Noctua Edition (performance mode)

61

59

020406080

Thermals Data

Product	Value
RTX 5080 Asus Noctua Edition (+400 core, +1000 mem, +10% PL)	62 Peak temperature (°C), 58 Average temperature (°C)
RTX 5080 Asus Noctua Edition (performance mode)	61 Peak temperature (°C), 59 Average temperature (°C)

Peak power (watts)

Average power (watts)

RTX 5080 Asus Noctua Edition (+400 core, +1000 mem, +10% PL)

405

382

RTX 5080 Asus Noctua Edition (performance mode)

370

352

0150300450600

Power Data

Product	Value
RTX 5080 Asus Noctua Edition (+400 core, +1000 mem, +10% PL)	405 Peak power (watts), 382 Average power (watts)
RTX 5080 Asus Noctua Edition (performance mode)	370 Peak power (watts), 352 Average power (watts)

RTX 5080 Asus Noctua Edition (+400 core, +1000 mem, +10% PL)

3195

RTX 5080 Asus Noctua Edition (performance mode)

2801

01,0002,0003,0004,000

Average clock speed (MHz)

GPU frequency Data

Product	Value
RTX 5080 Asus Noctua Edition (+400 core, +1000 mem, +10% PL)	3195
RTX 5080 Asus Noctua Edition (performance mode)	2801

The three G2 fans shut down when the card is under 50 °C. They start back up again at 55 °C. Even when this happens, they generate very little noise—not all that noticeable when I'm sat at my desk playing a game. A large fan can shift more air at slower speeds compared to smaller fan, and the Noctua Edition has very large fans compared to most. With low RPM and low noise, my test bench is actually too loud to pick up a specific reading for the Noctua card on my noise meter; the card operates quietly enough as to fade into the background hum of the PC.