Rumours continue to swirl around Intel's upcoming Nova Lake launch. The latest of these comes from noted leaker Jaykihn, who has hinted towards some exciting highlights among the desktop lineup. On X, the leaker wrote, "Nova Lake-H has a 12 Xe segment, and Nova Lake-S has a 12 Xe SKU. These are Xe3p."

Basically, this leak suggests there will be both a next-generation Nova Lake mobile CPU rocking 12 Xe cores as well as a desktop variant also sporting the same number of Intel graphics cores (though if that's true, I should probably be calling it an 'APU').

A high-performance iGPU was something sorely missing from the previous Arrow Lake generation of desktop chips, and having heard from our Dave what a 14 Xe core chip can do in the new MSI Claw handheld, it's exciting to think what a slightly upgraded version of Intel's GPU silicon could do in a wee little desktop chassis.

Anyway, drilling down a little more, the Xe 3P core is an upcoming iteration of Intel's microarchitecture. The discrete GPU tech was codenamed 'Celestial' for a time and was even meant to make its debut alongside Panther Lake. However, that chip went with what was essentially Xe2 'Plus Plus', and Jaykihn also claims what was Celestial 'was canned long ago'.

In other words, Xe 3P has been cooking for a minute. When it finally arrives in the form of the expected Crescent Island GPUs later this year, it'll be primarily geared towards handling AI inference by leveraging LPDDR5X memory, rather than the pricier data centre du jour HBM. It's entirely possible a powerful bit of kit like that could also turn its hand to gaming, but it's unlikely Intel would bring that chip to any gaming-focused cards.

As for a possible mobile APU, Nick recently wrote that a Nova Lake mobile chip with a beefy iGPU would be great news for handhelds, but that it would likely be destined for edge computing only.

But decent integrated graphics? In my desktop processor? That's exciting for sure, but how likely is it? Well, let's look at the leaker in question. Jaykihn claimed back in January the B70 Pro would arrive 'soon' and it did. Then in April, they leaked details about Intel's Core 300 laptop chips that also appeared to mostly bear out. Sure, a broken clock can be right twice a day, but this leaker is noted for having a pretty good track record as these things go.

Earlier this year, there was a rumour that Intel's next generation of desktop processors could consume up to 700 W of power. That was just for the so-called 'dual compute tile' models, with as many as 52 cores, but that claim now carries additional weight; one sharp-eyed leaker has noticed something about Nova Lake motherboards that points to the PL2 rating being way higher than it is now.

PL2 is the maximum sustained power limit Intel gives its chips under normal operation, and for all of Intel's current CPUs, it tops out at around 250 W. LC Tech Leaks on X gets credit (via BenchLife) for spotting something about a bare next-gen Z990 motherboard that all but confirms Nova Lake's highest PL2 being way bigger, specifically 220 W higher.

What we're looking at are the mounting points for three 8-pin power connectors, labelled 12V_2x4_1, 12V_2x4_2, and XTRA_12V. Current Intel motherboards such as MSI's MEG Z890 Ace have two 8-pin power connectors at the top of the motherboard to provide juice for the CPU, as the standard ATX connector is no longer suitable for supplying all of the current that a modern processor will draw.

That's been fairly standard for a few years now, so the only reason why a board would have an additional CPU power connector is if the default configuration is significantly larger than the typical PL2 figures that Intel's chips currently sport. For example, even the energy-sucking Core i9 14900K is only rated to 253 W (though it can, and does, use a lot more than that).

A single 8-pin EPS CPU power connector is good for something like 230 or so W, but can often handle way more than that, which is why motherboards with two connectors can give chips considerably more power. It's not hard to see why three connectors is proof of LC Tech Leak's claim that Nova Lake's PL2 is at least 474 W.

Now it's worth reiterating that all these hefty power claims are for the Nova Lake chips that are expected to support two compute tiles. The regular Core Ultra models, with eight P-cores and 16 E-cores, will just have one, and they're almost certainly going to have the same PL1/PL2 configuration as the current Arrow Lake processors, ie 125 and 250 W.

Intel is also expected to release Nova Lake variants with massive L3 caches to combat AMD's Ryzen X3D reign in the gaming market, some of which may require three power connectors, instead of the normal two. But I suspect that these too will be the dual compute tile ones, with up to 52 cores inside.

One thing that this all points to is that regardless of the chipset, the next generation of Intel motherboards will come in two tiers: one for single compute tile chips and the other for the twin tile CPUs. That's actually a good thing, because it should mean that PC gamers hoping to buy a cache-heavy Nova Lake processor can avoid paying even more for a specialised motherboard.

Certainly no gamer needs to be buying any of the dual compute tile models, as there are no games either out now or forthcoming that will need that many CPU cores. It will be a long time before eight primary cores become a significant bottleneck to performance, as there are hardly any cases when six isn't enough right now.

But as competition for AMD's Ryzen 9 and cheaper Threadripper chips, dual-tile Nova Lake could be the processor of choice for professionals. Assuming memory and storage prices aren't so high by then that we'll all be rendering and compiling via dreams and wishes.

With all-in-one liquid CPU coolers now so commonplace in gaming PCs, you might think there's nothing left for manufacturers to innovate with. But if the concept device that Cooler Master showcased at Computex is anything to go by, then we'll almost certainly see a raft of copy-cat designs on the market very soon, especially when Intel's Nova Lake chips make an appearance.

The idea behind it all is actually quite simple, but Cooler Master's execution looks really good: Take a 'standard' AIO cooler and slap a big fan on the pump/CPU block. When you look at the pictures, you'll realise that I'm not doing it any justice by describing it that way, but that's essentially what it is.

Here's what Cooler Master's PR manager, Brett Buren, told us about its G11M concept at the Computex stand: "[It's] kind of a hybrid between an air cooler and an AIO. So this is currently using our Atmos 2 radiator with the Master Fan A Gold. These are all aluminum fans, but there's a gold colorway."

Of course, a fan on the pump block isn't a new idea, and Arctic use it to great success with its Freezer III series. However, that system is used exclusively to provide a bit more airflow over the motherboard's VRMs; Cooler Master's design is more about chilling the water in the loop just a little bit more and filling the space inside a case, just as a traditional air cooler does.

"What's special about this part is there's basically an air cooler on top of the water block itself, so the water is being cooled in two places, one directly on top of the pump and one in the radiator," Buren told us. "Yeah, we had some comments like people were pointing out sometimes when you have an AIO, it looks kind of empty in your case."

I really like the look of it, so much so that if this is what the final design for the G11M looks like, I'll definitely be in line to pick one up. But there's another reason why I'm so interested in it: Cooler Master is targeting 400 W of thermal capacity for the device.

With the very biggest and best AIO coolers you can buy right now topping out at around the 350 W mark, depending on the fan speed, that's a fair increase in capacity. But aren't AMD and Intel's latest CPUs much lower than that? Yes, they are, but with Intel apparently planning to release a 52-core Nova Lake processor at some point in the near future, 400 W of cooling is likely to be a must.

Whatever happens on the processor front, I just want the sheer presence of Cooler Master's G11M glowering over everything else in my PC.

With Intel riding high at the moment, thanks to its excellent Panther Lake and Arrow Lake Refresh processors, any news about its forthcoming Nova Lake chips is worth paying attention to. Point in case: it's rumoured that Team Blue will release a 100% E-core mobile chip, with a big integrated GPU, with its next-generation architecture.

Admittedly, there's not an awful lot to go with Golden Pig Upgrade Pack's post on Weibo (via Wccftech), but what's there is interesting. The alleged processor will comprise eight E-cores and a 12 Xe-core GPU, and in the case of the latter, that's actually more than enough to run plenty of games at 1080p with a decent frame rate.

Intel's current E-cores in Arrow Lake are also surprisingly capable, so if we assume that the next (and possibly last) iteration of them is even better, but whatever the funky little chip is using, it could be an excellent low-power option for handheld gaming PCs.

Except there's a wee fly in the ointment in the fact that the leak suggests this chip will be purely for edge computing. Unlike Nvidia, which now classifies everything but AI and data centres as 'edge', Intel uses the term for any computational work that's essentially done where the data is generated. For example, robotic systems in a production line or autonomous driving.

When I visited Intel last year in Arizona, I spent a little time chatting with a development engineer about Panther Lake's NPU being used in AI edge applications. That part of the processor is great for low-power situations, but anything more demanding really needs a GPU behind it. We both agreed that an ideal chip for such situations would have a very low-power CPU section, but a really beefy iGPU.

Well, this is what we appear to be getting here, and I can't help but feel that it would actually be really good in a handheld. Alas, that's unlikely to ever happen, given its intended market, but you never know. One enterprising firm might create something experimental with such an edge chip and surprise us all.

The RAMpocalypse will almost certainly kill off any notion of such a venture, though, because prices for chips in the edge market are considerably higher than the domestic one. The combination of a unique, slightly odd processor that's expensive to buy, coupled with ludicrously pricey DRAM, would just make such a handheld unfeasibly costly. Bah, we never get anything nice these days.

You can always tell when an exciting new bit of PC tech is getting nearer to launch, because all the various leaks and rumours get increasingly more comprehensive and detailed. That's very much the case with the latest drop for Intel's Nova Lake chips, because now we have a look at what the full range of Core Ultra 400-series processors is going to look like.

Well, that's according to Videocardz, which says it's been shown a new roadmap for NVL-S, the next generation of desktop processors from Intel, which will be officially listed as Series 4, or Core Ultra 400S. As there has already been a wealth of leaks about the forthcoming chips, much of this information isn't especially new, but seeing all tied together and with snippets of details for the range of SKUs is interesting.

Just as with the current ARL-S (Arrow Lake) chips, and every other generation of Intel processors before it, Core Ultra 400S CPUs will come in a wide range of varieties for different segments: 125 W for general scenarios and gaming, plus 65 W and 35 W models for office and embedded applications.

PC gamers are most likely to be interested in the usual Core Ultra 9, 7, and 5 chips, and at first glance, it just seems to be the same as it has always been. The series leading 9-class models will sport nine Coyote Cove P-cores, 16 Arctic Wolf E-cores, and four Low Power E-cores (LPE).

Dropping down a tier to the 7-class will mean saying goodbye to four E-cores, and going further down the order to the 5-class results in losing two P-cores as well.

So, just as with Arrow Lake, then (though these don't have LPE-cores). However, where things get spicy are the bLLC variants, of which there is one apiece in the 9 and 7-class tiers. These are generally expected to be Intel's answer to AMD's X3D processors, as the bLLC is a monstrous slab of 'last level cache', purported to be up to 144 MB in size (for comparison, the 3D V-Cache slice in the Ryzen 7 9800X3D is 64 MB).

Whether a 7-class CPU gets that much additional L3 cache is anyone's guess, but I would reckon that Intel is looking to have a better separation between its Nova Lake SKUs than in Arrow Lake, so the full 144 MB might be reserved purely for Core Ultra 9 processors.

Now, so far, I've not said anything about the 'dual compute tile' models. That's because Videocardz's source for all of this information is apparently saying that Intel has yet to declare the brand nomenclature for the 44 and 52-core CPUs. This means these are unlikely to be sold as Core Ultra 9 chips, for example, and it's possible that Intel may choose to revive its X-series branding from 2017/2019.

The likes of the Core i9 10980XE Extreme Edition were huge, very expensive processors, with up to 18 cores and with quad-channel memory support. Purely for the HEDT (high-end desktop) market, such chips were for workstations and applications that required more threads than a 'normal' CPU could offer, but less than you'd get from a full-blown Xeon chip.

This is precisely what AMD's Threadripper chips are for, though as impressive as they are, they're really not for gaming at all.

Anyway, Videocardz also says that the leak confirms some other aspects about the Core Ultra 400-series, such as that the processors will still only have a maximum of 24 PCIe lanes available for graphics cards, SSDs, and other components (the same as Arrow Lake), but Nova Lake will support Thunderbolt 5, rather than 4, with up to 120 Gbps of transfer speed and 240 W of charging.

Assuming all of this is correct (and at this stage in the rumour mill, it almost certainly is), the only important pieces of information that are unknown are clock speeds, prices, and the all-important release date. Take your best guesses at those, folks.

Rumours around Intel's next-gen Nova Lake CPUs have multiplied rapidly in recent days. Highlights include an IPC performance advantage over AMD's upcoming Zen 6 chips, a core count upgrade for some models, and a massive APU variant to take on AMD's Strix Halo or its successor.

First up, those IPC claims. IPC refers to instructions per clock, and it's essentially a measure of how much work a CPU core does in each clock cycle. It's not a measure of overall performance, because IPC combines with clock speed to determine actual performance.

In other words, a processor's performance hinges on how much work a core does per clock cycle multiplied by the number of clock cycles per second. Anyway, established tech leaker HXL on X recently posted the following superficially cryptic message: "IPC: CYC>Zen6. Clock: Zen6>CYC".

For reference, "CYC" here refers to Coyote Cove, which is the generation of Performance core used by Nova Lake CPUs. With that in mind, the leaker is claiming that Nova Lake will beat AMD's Zen 6 when it comes to IPC, but Zen 6 will have a clockspeed advantage.

To be frank, with just those two data points, anything is possible. For instance, should AMD's clockspeed advantage be sufficient, it could still beat Intel regarding single-core performance. Exactly how high AMD will clock its next-gen Zen 6 cores isn't known, but some rumours point to figures in excess of 6 GHz thanks to the use of TSMC's most advanced N2P silicon.

Next up, Intel has apparently upgraded its mid-range Nova-Lake model from 42 cores to 44 cores. These variants have previously been said to be based on two Nova Lake compute tiles, each with 7 Performance, 12 Efficient and 4 Low Power Efficient cores. But the full-sized Nova Lake compute tile contains eight P cores. And now all eight are said to be enabled for this SKU according to another established X leaker, @jaykihn0.

Of course, the top Nova Lake chips have long been said to get 16 P cores, 32 E cores and four LPE cores, making for a grand total of 52 cores and a right old monster of a CPU. This new rumour does nothing to change that. What it does do is imply that whatever process Intel is using for Nova Lake, yields are probably decent.

After all, presumably the reason why Intel had previously disabled a P core in each tile on the previous 42-core model was to enable some chips with a dead P core to be used. If Intel has decided it doesn't need to do that, maybe yields are particularly good.

Another rumour involves a chip that is variously referred to as either Nova Lake AX or Razer Lake AX. Either way, what we're dealing with here is a monster APU that's said to use an absolutely huge socket known as LGA4326.

Indications of just such a socket have been found in shipping manifests on NBD Data attached to the NVL-AX code, which seems to be a pretty clear reference to Nova Lake AX. Some sources now refer to this chip as Razer Lake AX, and the implication here is that an earlier Nova Lake-based APU has been cancelled in favour of another APU using the newer Razer Lake architecture.

The most important aspect is that an LGA4326 implies an absolutely massive CPU socket with 4,326 contacts. @jaykihn0 on X says it's actually a BGA rather than LGA socket, but the point is that 4,326 contacts or pin outs is a massive step up from the 1,851 in Intel's current LGA1851 socket for desktop Arrow Lake CPUs.

Previously, Nova Lake AX was rumoured to run 28 CPU cores and 48 GPU cores, the latter based on Intel's upcoming Xe3P architecture. It's not known if the shift from Nova Lake AX to Razer Lake AX changes that, but given that this big APU is expected to be a chiplet design, it's quite possible that Intel could switch out the CPU tile but stick with its plans for the GPU tile.

On paper, a GPU tile with 48 Xe3P cores would be hugely powerful. Intel's current Arc B580 GPU has just 20 Xe2 cores. So, this rumoured APU has well over double the cores, and presumably each Xe3P core is more powerful than an Xe2 core. That said, @jaykihn0 nevertheless claims that there will be a version for laptops, which is intriguing, even if most observers think Razer Lake AX will be targeted at running local AI workloads rather than gaming. We'll just have to wait and see.

Taken overall, what we probably can conclude is that Intel's next-gen CPUs look like a much more exciting family of chips than anything Intel has produced in recent years. We're talking very powerful individual CPU cores combined with high core counts and maybe even the option of an APU variant with a properly powerful GPU.

As ever, rumours are just rumours until products are launched. But we've haven't seen rumours this exciting about upcoming Intel chips for ages.

For many years of back-and-forth between Intel and AMD, one thing has remained pretty certain: if you want platform longevity, you stick with the latter, because Intel changes sockets faster than I get through a pack of Leibniz butterkeks (very fast). But perhaps not for much longer, because Intel's Enthusiast Channel VP and GM, Robert Hallock, has just hinted he expects longer lasting motherboard sockets.

When asked by Club386 whether he sees Intel sockets supporting more CPU generations in future, he responded "I do. That’s it—I do."

Pretty unequivocal, if you ask me.

Admittedly this is just one man's opinion, but the man in question, Hallock, should have a ton of sway over the way things go given he's VP/GM of the 'enthusiast channel.' He was previously director of technical marketing at AMD, following other roles at the same company, so it's hardly surprising that he's taking socket longevity, traditionally a staple of AMD platforms, seriously.

Intel's existing Arrow Lake (Core Ultra 200S) and Arrow Lake Refresh desktop chips slot into motherboards with an LGA 1851 socket. It looks like the upcoming Nova Lake (Core Ultra 300S) will use a new LGA 1954 socket. This means anyone looking to upgrade from Arrow Lake to Nova Lake will need to get a new motherboard, too.

This is a familiar story for Intel CPUs. One of the reasons to opt for a current-gen AMD CPU is because AMD is expected to keep using its AM5 socket for next-gen Zen 6 CPUs, as it promised support for the socket through 2027 and beyond.

The AM4 socket lasted for four generations of CPUs. AM5 is the way to go if you can afford it, though; and the DDR5 memory to go with it.

With Intel, there are traditionally no such considerations. You get the CPU and future generations will more than likely require a complete platform shift. Maybe you get one more generation, if you're lucky. This might now change thanks to the new product management team:

"Every single one of us has built their own PC, games on that PC... there is a new engineering team for these gaming CPUs. And we are not ignorant of the feedback that comes in about our products."

If Intel can give the upcoming LGA 1954 socket a multi-generation lifespan, and get Nova Lake up and running using TSMC's N2 node, as we expect, Intel processors might stand a chance of being some of the best CPUs for gaming once again. It would certainly mean AMD shouldn't get complacent.

The Intel Nova Lake leaks are coming thick and fast today. Earlier, Nick reported on the apparently outsized power requirements of the dual-CPU-die variant of Nova Lake with 52 cores. Now come claimed details about the equally outsized proportions of the widely rumoured "bLLC" version of the Nova Lake CPU die with added cache memory.

In this context, "bLLC" stands for 'big last level cache' and indicates Intel's response to AMD's X3D V-cache CPUs, such as the Ryzen 7 9800X3D. The difference with Intel's approach is that instead of adding a separate slice of cache memory to the CPU die, as AMD does, Intel is reportedly planning to build the extra cache memory into the die itself.

If true, that will mean the bLLC CPU die will be substantially bigger than the non-bLLC die. And today, we have some apparently leaked numbers to put on that size differential.

According to X-based leaker HXL, the standard Nova Lake die comes in at 110 mm2. But the bLLC version clocks in at 150 mm2, quite a step up. For context, an AMD Zen 5 CPU die measures 71 mm2.

Making comparisons with AMD is complicated, though. The Nova Lake dies contain eight Performance and 16 Efficient cores, while the AMD die has just eight standard Zen 5 cores.

Meanwhile, the Nova Lake die is claimed to be built on TSMC's N2 silicon node, while Zen 5 is on TSMC N4, which is fully two generations of silicon older. If you're wondering why there's only talk of TSMC silicon, my understanding is that Intel may be farming out production of the larger CPU variants of Nova Lake to TSMC, while producing the smaller ones in-house on its new 18A node.

There is precedent for this approach in Intel's new Panther Lake CPUs. They use the Intel 3 node for the smaller version of the Xe3 iGPU, but TSMC N3 for the large iGPU. Anywho, the transistor density of TSMC's N2 node is dramatically higher than its N4 node.

Current rumours suggest that the big Nova Lake CPU die will contain 144 MB of bLLC cache. The AMD Ryzen 9800X3D currently has 64 MB of L3 memory in the V-cache slice, plus another 32 MB on-die. Of course, AMD will likely have its own upgraded Zen 6-based CPUs with even more V-cache around the time that Nova Lake is made available.

Anywho, Nova Lake is supposedly due out later this year, though for now it's not clear whether these purported bLLC models will be available from day one. On paper, they should be the most compelling Nova Lake models from a gaming perspective, as implied by the outstanding gaming performance of AMD's X3D chips.

Intel's past few series of desktop processors have somewhat of a bad reputation for being power-hungry beasts. Its current range of Arrow Lake chips has reversed this trend, fortunately, but with one top leaker claiming that the next generation, Nova Lake, will hit the 700 W mark, you'd be forgiven for thinking that Intel has gone backwards again. However, for PC gaming, you've got nothing to worry about.

That power figure is from kopite7kimi who posted on X that the "power consumption of a full-load NVL-K is over 700 watts." Since they're considered to be a reliable source of hardware leaks and rumours, there have been quite a few eyebrows raised over the purported power level.

The leaker goes on to clarify that this is for the "dual compute tile" version of Nova Lake, i.e. the one sporting 52 cores (two sets of eight P-cores and 16 E-cores, with an additional four low-power E-cores). If we assume that Intel really is making a dual tile version of Nova Lake, with that many cores, then 700 W is pretty much what I'd expect.

For example, the Core Ultra 9 285K (Arrow Lake, ARL-S) has eight P-cores and 16 E-cores, and when they're fully loaded to the hilt in Cinebench, that chip will happily consume up to 243 W of power. AMD's Ryzen 9 9950X, with 16 identical cores, tops 200 W in the same test.

If you scale the former up to a 52-core equivalent chip, you're going to be looking at over 480 W of power. Yes, that's still well short of 700 W, but Uniko's Hardware argues that the figure is most likely to be a PL4 value, and not the usual PL2. Intel defines these different power limits (PL1, PL2, etc) on the basis of time and usage, with the lowest being what the chip will happily consume all day long under load.

The higher ratings effectively become more burst-like, and most Intel processors will only sustain a PL2 figure for a certain amount of time before dropping to PL1. The highest level is almost instantaneous, a very short blast only. Lots of motherboards won't even let the processor run at that level, too.

Not that it matters for gaming anyway, because there isn't a game on the market, nor will there be any time soon, that will max out 16 P-cores and 32 E-cores. Heck, today's games barely stress a handful at a time, and the Core Ultra 9 285K only uses around 80 to 100 W of power in gaming.

That won't double with the biggest Nova Lake just because it has twice as many cores, as they're not going to be used all that much. You can see this when you compare a Ryzen 9 9950X3D to a Ryzen 7 9850X3D—the former only uses 9% more power than the latter, despite having 100% more cores.

It's a much larger difference (61%) when you compare the 9950X3D to the 9800X3D, but that's because the best gaming chip around is clocked very conservatively. The 9950X3D and 9850X3D are a bit more juiced, which is why they suck up more energy.

Besides, the vast majority of Intel gaming PCs that appear when Nova Lake chips hit the market later this year aren't going to be 52-core monstrosities. Some will for sure, but the majority will just stick to a single compute tile, because it will be far more affordable and still be perfectly fine for gaming.

I suspect that Intel will pitch the big Core Ultra 400 processors in the HEDT (high-end desktop) market, to compete against AMD's Threadripper monsters. That sector is generally a little less concerned about power consumption than gaming, as long as you're getting the right level of performance-per-watt.

So don't worry about Nova Lake's power figures, no matter how high they may seem. You've got far more important things to be concerned about, namely, how on Earth are gamers supposed to upgrade their rigs when DRAM and SSDs cost the same as fifteen organs on the black market?

With Intel's Nova Lake desktop CPUs expected to land later this year, we have finally got our first major leak of the motherboard chipsets to support them. The big takeaway is that they aren't all that different from the previous generation, but there now appear to be three of interest for prospective gamers.

Leaker Jaykihn took to X to post a chart with the specs of the Intel 900-series chipset, and if true, it suggests Intel has changed its naming nomenclature going forward. At the bottom end, we have the B960 and Z970, both of which offer 14 PCIe 4.0 lanes total.

Most of the specs of these two boards are similar, with two DMI Gen5 lanes to connect to the processor, one Thunderbolt 4 port, and an identical number of USB ports. The difference between the two is support for CPU overclocking—the Z970 supports it, the B960 does not, and neither support BCLK tweaking.

The B860, the current-gen Intel 800-series chipset, also has up to 14 PCIe 4.0 lanes, with the same number of SATA 3.0 and USB 2.0 ports. Notably, Nova Lake reportedly uses an LGA-1954 socket, which would make it not compatible with previous Intel motherboard chipsets.

The H chipset seems to be missing from the 900-series table, suggesting that Intel has cut its lowest-end chipset entirely. The H810 offered a lot fewer lanes than anything listed in the 900-series here, so perhaps it's gone entirely, and good riddance. Jaykihm even suggests there are "No H-series board at all" in a follow-up post. However, the H810 did launch later than high-end boards in the 800-series, so that's always a possibility again.

Historically, Intel used to use the Zx70 naming convention (like with the Z270), until AMD started doing the same with Ryzen-compatible motherboards. Intel shifted to Zx90 and Bx60 as a result, in what can only be described as a schoolyard game of 'my imaginary dinosaur is bigger than yours'. So it's a bit of a surprise that Intel appears to be going back to the Z970, which may look quite similar to any X970 motherboard from AMD.

Then, moving over to the Z990, which is the flagship chipset, we have 24 total PCIe lanes, which is the same as the Z890. Now, it uses four DMI Gen5 lanes, as opposed to eight maximum Gen4 lanes, but they offer a similar bandwidth, so there's not a distinct upgrade there. You're getting CPU OC, BCLK OC, and memory OC, as well as a large number of supported USB ports.

Previously, us gamers only really cared about the Z890 and B860 in the 800-series, there are now three worth considering: the B960, Z970, and Z990. If you are considering getting a Nova Lake chip, you will want to keep an eye on all three.

The Q970 and W980 are both chipsets intended for business users, explaining the presence of the Intel vPro+ Standard Manageability (though this is now called Intel vPro Essentials). This means their capabilities will be wasted in a gaming rig. Plus motherboards with them don't look anywhere near as cool.

As of the end of last month, Nova Lake CPUs were confirmed to be on track for a release in late 2026, with Intel saying 'We have important customers in both datacenter and client and that needs to be our priority'. Naturally, with the memory crisis ongoing, and the knock-on effects of that when it comes to hardware supply, we can never be fully confident with estimated release dates, but Intel's knocked it out of the park in Panther Lake testing, so it's all to play for. We just need Nova Lake to be worth the upgrade.

Nova Lake, the next-generation of Core processor, is set to launch at the end of the year. That's no surprise, but with the ongoing memory crisis making upgrading your processor and motherboard a very expensive ordeal, we had worried for both Intel and AMD's forthcoming generations.

In Intel's Q4 earnings call, the company's CEO, Lip-Bu Tan, confirmed the upcoming release window for Nova Lake:

"Along with our next-generation Nova Lake coming at the end of 2026. We now have a client roadmap that combines best-in-class performance with cost-optimized solutions giving me confidence that we are on the path of 45% market share and profitability in both notebooks and desktops over the next several years," Tan said (transcript via SeekingAlpha)

Nova Lake could be a big improvement over Arrow lake. I'm not getting my hopes up just yet, but it's rumoured to include a big last level cache (literally called bLLC) that will act as an analogue to 3D V-Cache on AMD Ryzen chips. Moreover, VP John Pitzer said he thought Intel would hold a leadership position with the launch of Nova Lake.

Though Intel will have to produce a very strong chip to both catch up to AMD's dominant X3D lineup and convince people to ditch their existing setups in favour of an Intel socket. Traditionally, Intel has not supported its sockets for very long—case and point, the LGA1851 socket for Arrow Lake is likely going to be dead and gone by Nova Lake.

Though as memory constraints hit our hobby hard, Intel confirms it is also constrained for supply. During the earnings call, one analyst from Morgan Stanley, Joseph Moore (fitting name), asks the company's CFO if they can shift more supply to meet large datacenter demand. Thankfully, Intel doesn't flatly oblige.

"We're absolutely constrained, Joe," says Zinsner. "So what we're doing within client we're focusing on the mid- and high-end and not as focused on the low-end. And then to the extent we have excess, we're pushing all of that into the datacenter space to meet that customer demand. And I think you'll see some share adjustments based on that because our primary focus is to our main customers.

"And obviously, we have important customers in the datacenter side. We have important customers in both datacenter and client and that needs to be our priority to get the limited supply we have to those customers."

So, that's more of a push for high-end Intel chips in client. Whether this affects Nova Lake is undetermined, however, it would have likely only ever been the high-performance K-series chips that launched at the end of this year anyways. That's the pattern throughout recent Intel history: K-series first, rest follow.

This plan of attack might go some ways to explaining why Intel has not announced a refresh of Arrow Lake yet. Certainly on the cards at points, according to rumours and benchmarks, if large volumes of chips aren't possible or necessary, and the high-end isn't going to sell due to performance concerns, perhaps the best course of action is to start fresh with Nova Lake later in the year instead.

There's a chance we'll see Arrow Lake relaunched and refreshed in the next few months, if only to please Intel's partners with new chips. Though I doubt there'll be much of interest for gamers until Nova Lake turns up. And even that may be a tough sell.

In the meantime, Panther Lake is shaping up very nicely for thin-and-light laptops, and perhaps a handheld too. So Intel has got some of its mojo back. Though it didn't arrive early enough to save Intel from a dip in revenue over Q4—$13.7 billion, down 4% year-on-year.

After a rough few years, Intel has clearly been hoping for better times, and it's fair to say that the official launch of Panther Lake at CES 2026 was a major step in the right direction. However, while all the attention was focused on the new mobile chips, the distinct lack of anything about an Arrow Lake refresh has cast doubt over whether it will actually happen or not. One hardware channel, though, is claiming that it definitely will.

That's according to Golden Pig Upgrade on Weibo, as reported by Videocardz, which says that the launch of the Core Ultra 200 Plus range (the rumoured name for the Arrow Lake refresh chips) will arrive in either March or April (via machine translation).

Evidence for the refresh itself comes in the form of entries in the Geekbench database. For example, this Core Ultra 9 290K Plus was uploaded only a few days ago, though it's possible that it's not genuine (i.e. it's actually a Core Ultra 9 285K result with a rejigged name).

Six months ago, I readily accepted the suggestion that Intel was planning on doing an Arrow Lake refresh, if only because it's done that sort of thing many times before in the past. Core 14th Gen processors are simply regurgitated Core 13th Gen chips, albeit with a bit more poke here and there.

The problem now is one of time. It's getting very late for Intel to be releasing a new range of desktop and mobile processors, because its next architecture, Nova Lake, is still on track for appearing later this year. That's not a rumour, that's literally what Intel has been saying.

So if Arrow Lake refresh does appear in March/April, that means it will only have a few months in the limelight before the next generation of processors comes crashing in to spoil the party. And if everything we've heard about Nova Lake comes to pass, then you'd be bonkers to buy a Core Ultra 200 Plus chip.

I think it would make more sense to just abandon the whole idea for the desktop market, but stick with it for the laptop sector, where Intel's Core Ultra 200HX chips perform very well. Tweaked versions of those would be nice to see in a gaming laptop.

One good reason not to bother with a desktop refresh is the fact that the main upgrade for the Arrow Lake refresh is supposed to offer native support for much faster DDR5 RAM. You know, the stuff that's now absurdly expensive. Why would you buy a new processor that's only a smidge faster than the current Core Ultra 200S family when you can't afford to pick up the memory kit to go with it?

Of course, AMD's Zen 6 and Nova Lake are both going to face that issue when they appear, because there's no clear sign as to when the DRAM crisis is going to end. But at least they're going to be properly faster and more capable than the current generation of CPUs, rather than being just a teeny bit faster. Well, I hope they are, at least.

At the end of the day, I don't think it really matters whether or not the Arrow Lake refresh appears or not. Fine if it does, equally fine if it doesn't, because either way, it's not going to sell anywhere near as well as Nova Lake will.

Now that we know Intel is reportedly getting ready to challenge AMD's X3D chips, gamers might start clamoring to buy Intel processors again. And if that happens, they'll need a cooler, because if there's one thing a good CPU can do, it's run hot.

To that end, Noctua has just quietly shared some good news for those who are already on an Intel platform: its coolers will still work with Intel's upcoming LGA1954 socket.

The company wrote on its FAQ page that the same coolers that have worked with the LGA1700 (Intel Alder Lake and Raptor Lake), as well as LGA1851 (Intel Arrow Lake), will remain compatible once the next gen rolls around.

Noctua products aren't exactly cheap, but they're powerful, which is why we deem the Noctua NH-D15 to be the best CPU air cooler. But it's one pricey beast. It costs $180 on Amazon right now, and Black Friday PC gaming deals seem to have had no impact on its price tag.

Intel Nova Lake really does sound exciting, so this is good news for gamers who are willing to give Intel another go. Arrow Lake's gaming performance was disappointing, but Intel seems to be making big strides toward ensuring that Nova Lake—aka Core Ultra 400 series—doesn't repeat those mistakes.

Rumoured to launch in late 2026, Nova Lake is said to be boosting the cache in a big way, which is something that helped AMD pull ahead in the first place. The top chip is said to sport up to 52 cores and 288 MB of combined cache. Oof, that's a lot.

On the other hand, Intel fans are once again being put through a test of patience, because even though the coolers may stay the same, the motherboards won't. Arrow Lake's LGA1851 socket will still welcome another refresh (Intel Core Ultra 300), but after that, it's game over, as Intel will move on to LGA1954.

Being able to at least save money on the cooler is a nice bonus. The other good part is that if Noctua's coolers will remain compatible with LGA1954, chances are that other manufacturers also won't gatekeep Nova Lake.

A CPU cooler is one of those things that can migrate with you from build to build, so this will be a neat little perk and a bit of savings when Intel Nova Lake hits the shelves around a year from now.

Intel is planning to sock it to AMD's dominant X3D CPUs with a 52-core version of its upcoming Nova Lake processor. Incredibly, it's claimed the chip will pack 288 MB of last-level vertical cache memory.

That's well over double the 128 MB of V-Cache currently offered by AMD's top X3D CPU, the AMD Ryzen 9 9950X3D. This news comes somewhat cryptically from one of the most reliable leakers in the business, kopite7kimi (via WCCFTech).

It's a slightly convoluted data trail. But to cut a long story short, kopite7kimi responded to a story discussing 144 MB verticle cache allocations, which is expected to be known as bLLC or big last level cache in Intel parlance. For certain versions of Intel's upcoming Nova Lake CPUs, kopite7kimi clarified that there will be models with two compute tiles and thus double the stacked cache.

So, according to this information, the top Intel Nova Lake CPU will have two compute tiles, each with eight Performance and 16 Efficient cores, plus 144 MB of vertical cache or bLLC. That makes for a total of 48 cores and 288 MB of Intel's equivalent to AMD V-Cache. The chips will also have another four low-power Efficient cores located in the SoC tile, making for a grand total of 52 cores (that core count total has appeared in a shipping manifest, too).

The reports claim that there will also be a variant with eight Performance and 12 Efficient cores and bLLC cache, albeit the amount of the latter isn't clear. Finally, there will be models with a single compute tile in both of the core and bLLC configurations mentioned. Overall, then, that makes for four Nova Lake CPUs with the bLLC option.

If all this is accurate, it makes for one heck of a broadside on AMD's X3D chips and a serious attempt to regain the gaming performance crown. Indeed, these plans are so aggressive, kopite7kimi claims that they will create problems for motherboard makers.

"The power supply design of the new platform will be a challenge for motherboard manufacturers," kopite7kimi said on X.

Of course, AMD will not be standing still as all this happens. Intel has confirmed that Nova Lake is due out at the end of 2026. But AMD has likewise said it will release its own next-gen Zen 6 CPUs in 2026.

All of which means that next year is shaping up to be a battle royale of the stacked cache gaming CPUs. Let's just hope the memory supply crisis has died down by the time these monster gaming CPUs arrive.

Intel's woes the last decade or so are well-enough documented. So, it's not a huge surprise to find a former senior Intel manager describing morale as being "in the toilet." But it hits hard, all the same.

A new feature on Fortune.com outlines Intel's current problems and provides a glimpse of the huge effort it's going to take to truly turn things around. Perhaps one of the most difficult challenges involves staff, both inside and outside the company.

On the one hand, staff inside Intel are said to be demoralised. It's a, “heads-down, push-through situation. That spark in people’s eyes, the desire to do this work, was not there,” said the same former Intel insider of the company's recent predicament.

At the same time, attracting new talent to what's seen as a sinking ship is going to be extremely difficult. Intel is no longer the top destination for the "brightest minds" in engineering. "Most people want to be part of a growth story, not a turnaround story," another former senior Intel manager is quoted as saying.

Fortune notes that Intel's latest CEO Lip Bu Tan has touted new engineering hires as evidence of the progress being made. But that reportedly hasn't landed well inside the company. "The message the employees got was: He’s bringing in people who are going to get paid way more than you are, and you’re losing your job," an Intel insider is quoted as commenting.

Anywho, the feature provides an interesting overview of how the company got into its current mess, chronicling Intel's shift away from being engineering-led to being finance-driven, recounting several major missteps, including the fateful decision not to make chips for Apple's then-new iPhone and Intel's more recent failure to get to grips with the AI industry.

It covers off the latest developments, too, including that stunning Intel-Nvidia deal and the scepticism with which some employees apparently view the US government's new 10% stake in the company.

It also touches on Lip Bu Tan's plan to emulate Nvidia with a "lean, fast moving" culture while also revealing how Tan's bombshell announcement that Intel will only continue to invest in its upcoming 14A node if there was sufficient customer interest came as a "shock to the system" for some insiders. That is a huge change from Intel's traditional "build it and they will come" approach to R&D.

All told, Intel is shaping up to be something of a case study in corporate malaise. Can Intel turn it around. Ironically, perhaps the most positive role model for Intel right now is also its arch rival, AMD.

As recently as 2015, financial analysts were predicting AMD would be bust by 2020. "We contend that Advanced Micro Devices Inc., (AMD) one of Silicon Valley’s semiconductor success stories, will file for bankruptcy by 2020," one analyst report claimed, before concluding, "time is missing for a successful transformation. With an uncompetitive product portfolio, several structural challenges coupled with poor positioning (resulting from an ineffective strategy), little hope is left for AMD given its worsening balance sheet. In our view, the conditions we outlined support our case that AMD is a true '0' and we do not foresee the company surviving beyond 2020."

That was just after a certain Lisa Su was appointed CEO of AMD. Fast forward just over a decade and AMD's market capitalisation or the company's total worth has grown over 100 fold, from around $2 billion to $267 billion at the time of writing.

Over the same period, Intel's market cap is slightly lower today at $172 billion than its $180 billion 2015 peak. And that's only thanks to a recent spike following the Nvidia deal announcement. Earlier this year, Intel's worth had sunk below $80 billion.

Still, one thing seems for sure. The Intel will be the subject of business classes for years to come. It's just not clear yet if the company will be a case study in how to turn around a large organisation in decline or a solemn warning of how a once great corporation can fall from financial grace.

If you've bought a gaming PC, or even thought about buying a gaming PC in the last year or so, there's a good chance you've seen them labelled as AI PCs. Turns out the desire for AI is lower than we might have thought, if recent Intel reports prove true.

According to Digitimes (machine translated), Intel is reportedly raising the prices of its elderly Raptor Lake chips by over 10%, and "industry analysts believe the price increase is likely related to TSMC and the lackluster performance of AI PCs." Raptor Lake CPUs, of course, do not have AI-accelerating NPUs.

The latest Lunar Lake chips have low power consumption and high efficiency, which makes them strong productivity chips (and good for AI) but pretty weak when it comes to gaming. According to the Digitimes report, "AI PC sales momentum is weaker than expected", so those running local AIs are less prevalent than otherwise expected. The same is presumably true on the desktop side with the latest Core Ultra chips as with laptop Lunar Lake ones.

Effectively, it appears to be seeing customers skip the latest chips in favour of the older ones. Digitimes doesn't note specific chips in its reports, but does clarify that chips selling for around $150 are climbing up to $170 or even $180. The report says, "Unfortunately, AI PCs are not the first choice for consumers."

Given its dwindling popularity compared to AMD, raising the prices of a relatively popular set of CPUs could appear a bit hasty, but if true, Intel could be trying to maximise profits going into the end of the year. With investment from the likes of Nvidia, and the potential of a deal from Apple, it likely wants to appear stronger financially going into future talks.

Raptor Lake had a pretty miserable launch, as Intel's 13th and 14th-Gen chips were hit with mass instability and crashes. A year and a half later, Intel was still releasing updates to fix problems with them.

Our testing of Raptor Lake chips like the Core i9 13900K or Core i9 14900K show a strong outing, considerably better than the likes of the Arrow Lake Core Ultra 9 285K or Core Ultra 5 245K that launched late last year. Intel has had a couple of rocky years, but there's solid productivity chops and a recognisable brand name under it all.

Arrow Lake is reportedly said to get a refresh next year, and it's no surprise that it's getting a new NPU specifically for AI workloads. If these recent reports prove true, there's a chance this refresh is a tad misguided.

Nova Lake is also reportedly launching late next year, and recent shipping manifests suggest the flagship SKU is packing a whopping 52 cores. Intel will need more than raw cores to catch up with AMD's rather dominant lead, though. Arrow Lake has seen performance fixes and price cuts since its launch, and Intel admits it 'fumbled the football', so a lot is riding on the future of Nova Lake. Perhaps the resurgence of sales for Raptor Lake is a sign that the focus will be less on AI. Though given the recently announced Intel-Nvidia partnership, somehow I think that might not happen.

Intel's been on a roll of late in all the wrong ways. Borked CPUs, failing fabs, haemorrhaging cash, run-ins with the President of the USA. The woes never end. Until next year, that is, and the launch of Nova Lake. That's right, peeps, when Intel's next-gen CPU architecture arrives at the end of 2026, Intel will ascend to a leadership position across the board on desktop.

To quote John Pitzer, Intel's Corporate Vice President of Corporate Planning & Investor Relations, precisely, "as Nova Lake comes out at the end of next year into 2027, I think we're going to have a leadership position across the board on desktop."

That is one heck of a bold claim, even for a chip that's rumoured to pack up to 52 cores. Of course, it will at least in part hinge on the quality of Intel's upcoming 18A process, the node formerly known as the one upon which Pat Gelsinger bet the entire company, but latterly more of a stepping stone back to profitability, with 14A being the node on which Intel's hopes to build chips for customers now depends on.

Unsurprisingly, Pitzer is bullish on 18A, too. He says Intel is planning to spend heavily on tooling up to produce lots of chips on 18A and return Intel to profitability. Famously, or you might say notoriously, both of Intel's latest CPU families, namely Arrow Lake and Lunar Lake, are largely manufactured by TSMC.

Intel will start the process of moving that production in-house with Panther Lake. However, Panther Lake is a mobile CPU and is also aimed at more power-efficient laptops as opposed to larger desktop-replacement rigs. More to the point, it's thought only certain versions of Panther Lake will have CPU dies made by Intel. The bulk of production may well end up being handled by Taiwanese chip foundry TSMC.

In other words, Panther Lake will only entail fairly limited quantities of Intel's 18A silicon. That changes with Nova Lake. "Nova Lake itself being both a notebook and a desktop part has pretty meaningful implications for the amount of wafer starts that we need on 18A," Pitzer says.

And that should mean improved profitability because it means Intel won't have to pay those pricey TSMC wafer fees. "As you look at it through the lens of Intel Foundry, the move from Intel 7 to Intel 18A, ASPs per wafer will go up three times faster than their cost. And so just driving more volume through the fab on 18A is a pretty profitable dynamic for Intel Foundry," Pitzer.

Morever, Pitzer says Intel can get back to making money just by executing on its own products, it doesn't absolutely need to win big foundry customers. "We don't need to see a lot of external foundry revenue to breakeven exiting 2027," he says.

The question is, can Intel actually deliver on its own products? If you examine Intel's recent CPU families, it's not altogether promising. The Raptor Lake generation has turned out to suffer from major bugs, the Meteor Lake mobile architecture was underwhelming, and Intel's latest desktop chips, known as Arrow Lake arrived half baked and even after a little tweaking remain well behind AMD by most estimates.

The only unambiguous exception to all that has been Lunar Lake, which is a decent low-power mobile CPU, but according to Intel itself, isn't a money spinner on account of being made mostly by TSMC and having integrated memory, which limits configuration options for laptop makers.

We'll get an initial feel for whether Intel is getting back on track with Panther Lake at the end of this year, which will be our first taste of the 18A process. But that's another low-power CPU. So, it will really be Nova Lake at the end of 2026; that's the real test.

Nova Lake will span the whole gambit from lower-power laptops to high-performance desktops and is shaping up to be an absolutely vital processor family for Intel. Nova Lake simply has to be at least competitive with AMD if Intel is going to turn things around.

Like most gaming PCs, yours will probably be using overclocked RAM sticks. You might not think it is, but every time you enable XMP or EXPO, that's precisely what you're doing. So, fans of Intel's latest CPUs might be excited to learn that Asus has just announced that its 800-series boards now support DDR5-7200 without overclocking.

And it's the most casual and simple of announcements, as you like, with the post on X (via Videocardz) simply stating:

If you look carefully at the pictures in the post, you'll see some screenshots of the free software tool CPU-z showing the RAM speeds and timings of a kit in an Asus Prime Z890-P motherboard. Zoom into the section for the SPD tab and you'll see several columns labelled JEDEC #14 through to JEDEC #17.

JEDEC (Joint Electron Device Engineering Council) is a consortium of semiconductor and electronics companies that work together to agree on standards for all things PC-related. In this case, it's about RAM speeds and timings. Anything labelled Jedec is basically classed as 'not overclocked'—ie the RAM and CPU are all good to work at those ratings by default.

Intel's Core Ultra 200S processors support DDR5 running up to a JEDEC-approved 6,400 MT/s, though they will happily cope with faster RAM than this. I say happily, but it's not 100% guaranteed that your Core Ultra CPU will handle DDR5-8400, for example, because that's only achievable by officially overclocking the RAM and the processor's memory controllers.

But what Asus is saying is that if you use one of its 800-series motherboards, you'll be good for hitting 7,200 MT/s sans overclocking. Which puzzles me a bit—why didn't Intel say that its CPUs were capable of this during the launch of Arrow Lake? It could have said something along the lines of 'Our processors natively support up to 7,200 MT/s and our launch partners will have products to comply with this in the near future. For now, you get 6,400 MT/s.'

One possible explanation is that Intel is actually keeping this for the expected Arrow Lake Refresh. It's generally thought that, just like the last-gen 14th Gen Core lineup, the next release of Intel processors will be nothing more than clock-bumped Arrow Lake chips. Having a higher default memory controller clock would naturally be part of that.

If so, then does that mean Asus has jumped the gun here and given us a wee snippet about what Intel has planned for us while we wait for the goliath-sized Nova Lake processors?

I've contacted Asus to ask about what the post means for current 800-series owners and whether they'll need to grab a BIOS update to have this JEDEC DDR5-7200 or not, but at the time of writing, it hasn't got back to me yet. We'll update this story as soon as we hear more.

As for using ultra-fast RAM with your Core Ultra 200S chip, well, I tested a whole host of games and applications with speeds up to 8,200 MT/s and it didn't make a lot of difference, especially in gaming. However, that was before all the various microcode, BIOS, and Windows updates we've seen since Arrow Lake launched last October.

Even so, I'm not confident that 7,200 MT/s will have any noticeable effect on your games' performance, as Arrow Lake's biggest weakness is poor RAM latencies. Higher clock speeds help mitigate this, but to completely counter it, you need a big boost in clocks all around and not just on the RAM and memory controller. Oh, and an improved cache system.

In other words, what we need is just a better processor altogether. One called Nova Lake, perhaps?

Sometimes, bigger isn't always strictly better. More cores can equate to a higher power draw, and there's a fine balancing act when building a PC between getting parts that run nice, and ones that don't require a nuclear reactor in your case to work. Intel's upcoming Nova Lake chips seem to be rocking a lot of cores, but whether or not that will be worth it will depend on more than raw figures.

According to @meng59739449, a leaker on X, a shipping manifest included mention of two different Intel Nova Lake chips. The first is a 28-core chip with eight Performance cores, 16 Efficient cores, and four Low-Power E-cores. The second, perhaps the most interesting one, has a pretty chunky 52 cores, with 16 being Performance cores, 32 being Efficient cores, and 4 being Low-Power cores.

This all lines up with rumours from June suggesting the same number of cores. For context on that figure, the Intel Core Ultra 9 285K has eight Performance cores and 16 Efficient cores. That means the top model has more than double that amount. That is an impressive jump in raw performance.

This new information was shortly followed by the confirmation of an Engineering Sample (ES) arriving at the end of the year. Given that Nova Lake was previously rumoured to launch in 2026, this lines up with previous reports. Notably, the manifest shows that the new Nova Lake models use an LGA-1954 socket, which is entirely new.

This means, if you upgraded to an LGA-1851 motherboard when it launched in the latter half of 2024, you will now have to go out and buy a new one should you want Nova Lake when it is expected to arrive in 2026.

Nova Lake reportedly operates on a mixture of TSMC's N2 node, as well as Intel's own 18A one. There's a lot riding on 18A, and just this week, Intel's CEO admitted that it 'fumbled the football' with the launch of its Arrow Lake CPUs. Intel claims Nova Lake "is a more complete set of SKUs. It does address the high-end desktop market. And so we would expect that we will improve our position next year. So all in all, I actually feel pretty good about the client."

18A is reported to achieve a peak transistor density around 185 MTr/mm2, which is still behind the likes of TSMC and behind recent 2n rumours from Japanese firm Raipidus. It is supposedly built to optimise performance and efficiency and comes with backside power delivery (BSPDN), which should increase performance-per-watt.

However, Intel's biggest weakness right now isn't the number of cores, as Nick points out, it's instead cache latency. Intel is reportedly introducing a gaming-friendly cache memory tile to rival AMD's rather brilliant X3D chips, but adding more cores can contribute to core-to-core cache access time.

More cores will certainly contribute to a CPU being great in multi-threaded applications, making them particularly good as productivity machines, but for them to compete in the gaming market, they need low-latency data access. This is done through a strong cache structure.

Intel reportedly stated it was looking for "leadership gaming performance" at the end of June, but shoving more cores into its chips isn't how you beat AMD in itself. It will need a smart and optimised bus structure to go with it. Hopefully, the larger range of SKUs will allow some space for a high-end productivity chip and a gaming one alongside it. The next year will be very important for Intel.

It should come as no surprise that the biggest chip companies in the world spend billions of dollars every year on research and development. To stay ahead of the competition and improve revenues, new architectures, systems, and software stacks need to be designed, built, and tested. What might raise a few eyebrows is just how much money Intel spent last year on R&D, especially compared to AMD, Nvidia, and the rest of the semiconductor world.

An analysis of research and development expenditure by TechInsights was reported by Korea JoongAng Daily, but you can get the numbers yourself by pulling up each company's 2024 financial results. For example, AMD declared that it spent $6.456 billion last year (pdf, page 1) on R&D, whereas Nvidia forked out $12.914 billion. It's worth noting that Nvidia's financial statements are numbered one year ahead of the actual period (FY 2026 is 2025 and so on).

Anyway, those figures pale in comparison to how much cash Intel burned through in 2024 to research and develop chip, fabrication technologies, software, and all kinds of tech stuff—a staggering $16.546 billion (pdf, page 25). That's 28% more than Nvidia and a frankly unbelievable 156% more than AMD.

The nearest non-US semiconductor firm is Samsung Electronics, which spent a reported $9.5 billion on R&D. That would place third, comfortably ahead of AMD, and it strongly suggests that if you have your own foundries for making chips, you need to spend a lot of cash on finding ways to make better processors.

When you compare the R&D expenditures to the amount of revenue, though, the story takes on a very different look. Intel spent 31% of its net revenue, and 26% for AMD, but Nvidia and Samsung got by on just 10% and 4%, respectively. Team Green's percentage looks very low, but only because it made so much money through its AI products.

It's interesting to see that AMD is committed to spending a healthy chunk of its income on R&D. Unlike Intel, it doesn't really have to worry about chip manufacturing and process nodes, though it will spend some cash in this area. That means the vast majority of the $6.456 billion it spent in 2024 was for making better CPUs, GPUs, and other accelerators, as well as the software to support them all.

Intel's product portfolio is perhaps the largest of the US semiconductor gang, though to cut losses, it's been dropping big projects and trimming divisions left, right, and centre. Whether that means its 2025 R&D costs are significantly lower is anyone's guess right now, but I suspect that the current CEO, Lip-Bu Tan, will probably want to trim it down a fair bit.

That means by the end of 2025, there's a very good chance that Nvidia will be top of the R&D spending table, partly because it will be concerned about losing ground in the AI market to cheaper competition, and partly because it has vast amounts of cold, hard cash to do so. Intel and AMD would probably like to spend more, but can't because they're either losing money as things currently stand or because their revenue stream just won't support it.

What none of them will want to do, though, is spend less on R&D for their core products: CPUs in the case of AMD and Intel (plus Samsung, to a lesser extent) and GPUs for Nvidia. These products form the backbone of their income, and with Team Red and Green laughing all the way to the bank because of how well their Ryzen and GeForce/AI products are doing, they'll certainly keep splashing the research moolah.

As for Intel, while Arrow Lake hasn't been an absolute disaster, it was certainly a 'football fumble' and I'm sure Team Blue's bean counters will be hoping that all the cash it's spent on developing Nova Lake transpires to have been worth every cent.

Despite receiving numerous updates and enhancements since its rough launch in October 2024, Intel's Arrow Lake desktop CPUs have not been especially popular, and for multiple reasons. At a recent technology conference, the chip company openly admitted that it had "kind of fumbled the football … particularly [the] high performance desktop side", but also promised that "Nova Lake is a more complete set of SKUs."

These specific comments were said by Dave Zinsner, Intel's Chief Financial Officer, in an interview at the 2025 Deutsche Bank’s 2025 Technology Conference (via Wccftech). In addition to discussing developments in its financials and manufacturing, Zinsner was also asked about Intel's current CPU product line, specifically what was working well and what needed improvement.

Opening with its mobile products, Zinsner was upbeat: "I think on notebook, we’re in pretty good shape. We introduced Lunar Lake this year. It’s in the process of gaining adoption. We’re expecting this quarter to be a pretty good quarter for Lunar Lake. So notebook, I think is good, solid, share is solid."

I can only assume that Intel's CFO made a slip of the tongue because Lunar Lake, a potential little chip exclusively for low-powered platforms, was actually launched in May 2024, though the first processors didn't start rolling out until September. Core 200V chips are indeed fine chips—suitable for slim laptops and handheld gaming PCs—but I'm not convinced that the product is gaining as much adoption as Intel would like, as the embedded DRAM makes them more expensive than the competition.

However, Zinsner was a lot more frank about Arrow Lake, its multi-tiled processor for desktop PCs. "As you know, we kind of fumbled the football on the desktop side, particularly high performance desktop side. So we're as you kind of look at share on a dollar basis versus a unit basis, we don't perform as well and it's mostly because of this high end desktop business that we didn't have a good offering this year."

What he's referring to is the fact that the Core Ultra 9 285K just isn't good enough when compared to what AMD has on offer. For pure gaming, the Ryzen 7 9800X3D isn't just faster, it's also a lot cheaper. It's a similar situation for productivity and content creation, as the Ryzen 9 9950X is roughly on par with the 285K but undercuts it on price.

It doesn't help that the Arrow Lake SKU range is rather poor, with the Core Ultra 5 245K being very underwhelming, and there has been no direct replacement for the Core i3, so far. Only the Core Ultra 7 265K is truly worth considering, as it's not that much slower than the 285K and it's wildly cheaper.

While Zinsner didn't offer any explanation for the 'football fumbling', he did go on to say that what Intel has planned next will make things all hunky-dory again, albeit rather uncommittedly so.

"Nova Lake, which is the next product, is a more complete set of SKUs. It does address the high end desktop market. And so we would expect that we will improve our position next year. So all in all, I actually feel pretty good about the client."

Nova Lake will be another multi-tiled processor, and over the past few months, we've seen all kinds of rumours about enormous P-core and E-core configurations and cache enhancements to go up against AMD's 3D V-Cache. Just as with Arrow Lake, most of the tiles will be manufactured by TSMC, though the main compute tile is expected to be straight from Intel's own foundries.

The biggest problem that Nova Lake has to face, though, is that not only does it need to address the performance and price deficit between Arrow Lake and AMD's Zen 5, but it will also need to address the likelihood that Zen 6 will be around at the same time. In other words, the next generation of Core Ultra processors will need to be significantly better than the current series, if Intel is to claw back ground lost to the Ryzen 9000-series.

Naturally, the chief financial officer of any large company is never going to be downbeat about future products, and I expect every senior leader to talk up Nova Lake at every opportunity. However, Intel has far more on the line than just desktop market share. It's been steadily losing market share to AMD in the world of servers for several years, and it's all but given up in the AI sector, as it doesn't have anything to combat AMD and Nvidia's massive superchips.

Even if Nova Lake turns out to be extremely good, so much so that Arrow Lake is instantly forgotten and PC gamers around the world abandon Ryzen X3D chips in favour of the new Team Blue chips, the big money lies in sectors other than desktop computing. And cold, hard cash is something that Intel desperately needs right now to pay for its never-ending foundry losses.

AMD's Strix Halo APU with its outsized integrated graphics is of debatable benefit for PC gaming (more on that in a moment). But apparently Intel still wants to take on AMD in the uber-powerful APU market with a variant of its upcoming Nova Lake generation of chips.

So say rumours emerging today. It started with a relatively unassuming X post referring to "Nova Lake-AX" and describing it as "preliminary". WCCFTech has since fleshed that out, claiming that Nova Lake-HX chips will be, "halo-class enthusiast CPUs, which would compete with AMD and Apple SoCs such as the Strix Halo and M4 lineup."

As for specs, it's said Nova Lake-AX will pack two CPU tiles, each with eight P cores and 16 E cores, plus four ultra-low power LP-E cores located on a separate "low power" island and thus a grand total of 52 CPU cores. Of course, the really interesting bit would be the iGPU.

That's claimed to be based on Intel's next-gen Celestial graphics architecture, otherwise known as XE3. Intel's current most powerful iGPU is in Lunar Lake (pictured above). It uses XE2 or Battlemage technology and sports eight execution units or EUs.

WCCF speculates that the GPU tile in Nova Lake-AX could include as many as 20 to 24 XE3-spec EUs, which would be a huge upgrade. For context, AMD's Strix Halo APU has 40 CUs or compute units. They're not directly comparable to an Intel EU. However, AMD's mainstream APU, Strix Point, tops out at 16 CUs.

In other words, Strix Halo has well over double the CU count of Strix Point. And if you apply similar scaling to Intel's current top iGPU, you get to 20-plus EUs. Again, nothing is confirmed, but it all superficially scans.

Of course, as I alluded to up front, this Nova Lake-AX chip may not necessarily shape up as a great gaming solution. AMD's Strix Halo, while very powerful for an APU, isn't ideal for gaming.

Primarily, that's because it's very expensive for an APU. AMD has given Strix Halo 16 CPU cores. If what you wanted was an affordable APU alternative to, say, Nvidia's discrete mobile GPUs for laptops, you wouldn't make the CPU very expensive with flagship specs. An eight-core CPU would be plenty. After all, our favourite desktop CPU, the AMD Ryzen 7 9800X3D, only has eight cores.

Instead, Strix Halo is much more interesting as a chip for creatives and people wanting to run small AI models locally on a laptop. And that is probably what this Nova Lake-AX APU will also be best at, rather than gaming. At least from a bang-per-buck perspective, you'll still likely be better off with a discrete GPU.

It hasn't exactly been the best year for Intel. The chip giant has faced major struggles recently with chips flopping and fab partners seemingly nowhere to be found. Even Bill Gates has lamented the current state of the company, but maybe the upcoming Nova Lake-S CPUs, will turn all that around.

We've seen a lot of rumours around the Nova Lake launch, and while some make more sense than others one has likely been confirmed. According to SemiAccurate (Via TechPowerUp) the next generation CPUs have been taped out on TSMC's N2 node in fabrication centres in Taiwan.

Well, kind of. SemiAccurate is actually pretty slim on the details.

"Intel taped out a major product a few weeks ago, a little late but they got there. SemiAccurate took longer than usual to confirm this one but we finally did.

"Sorry, no hints this time." reads the website. I guess that is semi-accurate.

Assuming this is the Nova Lake chipset, this likely confirms the hybrid nature of the tech, using a mix of both the N2 node as well as Intel's highly anticipated 18A node. It could be Intel has gone with a mix due to a lack of faith in the 18A, or just to maximise manufacturing capabilities with more fallback on the established N2 nodes.

The chase of Intel's 18A node is an interesting one. We expected to see these chips launch in PCs but it seems Intel just didn't have the faith to launch them. Of course with these CPUs only being taped out now, these processors could be a fair way off, with at least a few months being the norm from taped to manufacturing, let alone retail availability.

We'll likely be waiting at least a year to get our hands on these, but with delays around the 18A chipset maybe it'll be even longer. Hopefully we'll see the 18A powered Nova Lake chips before the end of 2026.

In terms of performance, the numbers look good. The CPUs are reportedly delivering up to 25% faster processing with up to 36% more efficiency than before. Intel's new chips are also rumoured to have over double the cores of its previous Arrow Lake's offerings. And though we'll probably be waiting another year for them, a new chip being taped down is always cause for celebration.

Despite not being expected to appear until 2026, rumours and leaks about Intel's next generation of desktop CPUs continue to come thick and fast. The latest snippet claims that Nova Lake will have "leadership gaming performance" coming from more than 10% better single-threaded (ST) and 60% higher multi-threaded (MT) performance gains.

However, it's not a particularly comprehensive leak, just a snapshot of what looks like an internal Intel document, posted by X user Game.Keeps.Loading (via 3DCenter). First of all, it doesn't explicitly state that it's about Nova Lake, the successor to Arrow Lake, but the fact that "New Low-Power Island" is a feature pretty much points to it being about the next generation of desktop CPUs. This is just a cluster of low-power E-cores, used for background tasks, that first appeared in Meteor Lake but not in any desktop chip so far.

Other than that, there are just two additional statements: "Leadership Gaming Performance" and ">1.1x higher ST and 1.6x MT performance". The former is standard marketing fluff, and there's nothing to really note here, other than Intel is clearly trying to get one over AMD in the gaming market. The Ryzen 7 9800X3D is so much better than the Core Ultra 9 285K in games that it's obvious Intel can't ignore it.

The single- and multi-threaded (ST/MT) performance gains are another matter, though. An uplift of 10% for ST is almost certainly referring to the P-cores, and while that might not sound like much of a boost, it's par for the course with CPU architectural changes these days. After all, AMD claimed that Zen 5 had a 16% better average IPC (instructions per clock) than Zen 4, but games didn't show anything like that at launch.

However, 60% better multi-threaded performance is an odd one, partly because there's no indication what the comparison is with and partly because it's been rumoured that the Nova Lake range will top out with a 16 P-core, 32 E-core monster. If one assumes that Intel is comparing that chip to the 285K (8 P-cores, 16 E-cores), then clock-for-clock, you're looking at a 100% uplift in MT performance.

So where's the other 40%? It's possible that a 16P+32E monster CPU just can't maintain anywhere near its maximum clock speeds when fully loaded with threads, but if it's dropping that much, it's a bit worrying. One might argue that Intel isn't planning on launching with a 48-core beast and instead will start off with an 8P+32E or a 12P+24E chip at best.

A Nova Lake processor with 36 or 40 threads would have an MT uplift over the 285K of 50% to 67%, clock-for-clock, which is more in line with the 60% claim in the leak, but even so, the numbers just don't make a lot of sense right now. Especially that the one thing Intel chips don't need more of is cores.

One concern I have with Intel throwing a mountain of cores into its next desktop CPUs is how they access and share the last level of cache or LLC (currently L3, in the case of Arrow Lake). Each P-core and E-core cluster has its own slice of LLC, but they can access other slices via a ring bus. The more cores/clusters one adds to the ring, the more ring stops need to be present, and the higher the latency one gets with core-to-core cache accesses.

Cache latency is currently Intel's biggest weakness against AMD's Zen 5, especially when 3D V-Cache is brought into play. While it's being claimed that Intel has something planned to counter the dominance of X3D chips in gaming, just adding more cache isn't the solution—the latency really needs to be cut right down.

Intel's older Raptor Lake chips (i.e. Core 13th/14th Gen) are better than Arrow Lake (Core Ultra 200S) in games, because they can clock much higher, especially the ring bus. The downside is that they will suck up as much power as you can throw at them in the process and Intel had to do something about that, hence why Arrow Lake's clock speeds and power consumption are far more conservative.

I've spent a lot of time over the year messing about with voltages, clocks, and various timings with a Core Ultra 9 285K and Ultra 7 265K to try and improve how well the chips can fling data around. I can get some pretty decent throughput figures, but the access latency for L3 never really improves enough to make any headway against a Ryzen 7 9800X3D.

Nova Lake could be an absolute behemoth in multi-threaded applications, especially those used in the professional sector, but games are still more than happy with 12 to 16 threads. That's because all the big releases are developed for consoles, which only offer that many threads to developers. What they want is a really good cache structure, with fast, low-latency data access.

If Intel's move to aim for "leadership gaming performance" consists entirely of just shoehorning a buckload of cores into its next generation of desktop processors, I suspect that AMD won't have much to worry about.

AMD launched its first X3D CPU with 3D V-Cache back in April 2022 in the Ryzen 7 5800X3D. Since then, AMD's X3D chips have pretty much been the weapon of choice for well-funded gamers. Where, you might ask, for art thou, Intel? According to the latest rumour, it might finally have an answer to X3D when it launches its next-gen Nova Lake CPUs in 2026, or perhaps early 2027.

X user Haze2K1 has posted some specifications of purported Nova Lake CPU models and one detail stands out (via Club386). Along with listing core counts and TDP, both CPU models allegedly get something called "bLLC".

That, supposedly, refers to something known as big Last Line Cache. And it's very much analogous to AMD's 3D V-Cache. Indeed, we know for sure Intel has such a technology, because it's incorporated in the new Clearwater Forest generation of Xeon server CPUs. But Intel has previously said it has no immediate plans to bring that technology to the desktop.

The specifics of bLLC, at least as it pertains to that Xeon chip, entails what Intel calls Local Cache integrated into the base tile. The base tile is a chiplet in a modern Intel CPU package that sits beneath the active tiles, largely serving as an interconnect. At least, that's it's job in multi-die CPUs available today from Intel such as Lunar Lake and Arrow Lake.

By adding cache to the base tile, Intel would end up with a similar broad approach to AMD's latest X3D CPUs. AMD's first two generations of X3D chips had the V-Cache attached to the top of the CPU dies, which was not optimal for thermal performance and therefore clock speed.

However, for its third generation X3D CPUs, including the Ryzen 7 9800X3D, the V-Cache was moved below the CPU chiplets, essentially banishing the clock speed disadvantage of previous X3D CPU models and turning the latest generation into absolute gaming beasts.

But before we all assume Intel's Nova Lake CPUs will therefore be killer for gaming, simply adding cache memory isn't a guarantee of better frame rates. The whole point of cache memory is to improve memory latency by reducing the need to go out over the main memory bus to the system RAM. And that, in turn, improves performance, especially in games.

But for that cache to be really effective, it has to offer very low latency itself, which isn't a given. AMD's memory bus and cache architecture is highly optimised for low latency, but it's yet to be seen if Nova Lake will match AMD by that measure.

AMD's has a super fast point-to-point internal interconnect for its CPUs, while Intel currently uses a ring bus to connect cores, graphics and higer-level cache memory. That ring interconnect is generally slower and suffers higher latency.

By some measures, that gives Intel's Arrow Lake 512 GB/s of internal bandwidth to the 2.5 TB/s of AMD's equivalent technology. Anyway, the point is that Intel will need to do more than merely stuff some cache into the base tile to match the impact of AMD's 3D V-Cache in games. In other words, watch this space.

According to a detailed post on X, Intel's next-gen Nova Lake desktop CPU will be getting over double the cores of its existing Arrow Lake chips. The top Core Ultra 9 model allegedly packs a staggering 52 cores. But it's arguably the mid-range Ultra 5 that's most interesting given it boasts more cores in every category than Intel's incumbent top desktop processor.

The current Intel Core Ultra 9 285K has eight Performance and 16 Efficient cores. However, according to the X post, there will be a Nova Lake Core Ultra 5 model with eight Performance, 16 Efficient and another four Low Power Efficient cores.

Meanwhile, the top Core Ultra 9 model crams in 16 Performance, 32 Efficient and four Low Power Efficient cores for that grand total of 52 cores. Even the very lowest end Nova Lake gets 12 cores, with a 4P, 4E and 4 LP-E split.

If true, Nova Lake will be the biggest jump in raw CPU performance from Intel in some time. Intel's desktop chips have topped out at eight Performance cores since the Alder Lake generation launched back in late 2021.

That generation also offered eight Efficient cores. But while the Raptor Lake follow-up boosted the E-Core count to 16 a year later, Intel hasn't increased core counts since. Indeed, Intel actually deprecated the total thread count when Arrow Lake arrived in October last year on account of removing support for HyperThreading, which enables Performance cores to support two software threads in parallel when present.

Anyway, if these core counts are correct, the multi-threading performance of Nova Lake will be pretty epic. If Nova Lake also brings improved IPC from its Performance cores, thought to be codenamed Coyote Cove, and Efficient cores, codenamed Arctic Wolf, then the overall performance uptick could be spectacular.

As for how this compares with AMD's future plans, it isn't totally clear. Various rumours point to anywhere from 12-core to 32-core chiplets in AMD's next-gen CPU plans using the upcoming Zen 6 architecture. The former would probably mean a 24-core top desktop CPU, the latter as many as 64 cores given AMD's top desktop CPU conventionally has two CPU core chiplets.

However, the 32-core chiplet is probably based on the Zen 6c architecture with compact cores with the full Zen 6 chiplet topping out at either 12 or 16 cores. That would give total core counts of 24 and 32 respectively. With multithreading, you'd be looking at 48 or 64 threads.

If you take a pessimistic view, that's 48 threads from 24 full fat Zen 6 cores versus 52 mixed cores from Intel. Game on. However you slice it, it certainly looks like desktop PCs will benefit from a very meaty upgrade when Nova Lake and Zen 6 arrive.

As for exactly when that will happen, we'd bank on late 2026 for Nova Lake in terms of a launch date with early 2027 a more realistic target for widespread availability. It's not yet clear what production node Intel will use for Nova Lake, with Intel's own 18A and 14A nodes, along with TSMC N2 all mooted as possibilities by various rumours.

AMD's Zen 6, meanwhile, may be based on TSMC's N2 node when it arrives, likely in the second half of 2026. AMD has confirmed that the server variant of Zen 6 will definitely use TSMC's next-gen N2 node, which heavily implies, though doesn't absolutely guarantee, that Zen 6 for PCs will use the same technology.

Anywho, the latter half of 2026 is certainly shaping up to be pretty exciting for the PC in terms of new CPUs.

Intel's numerous woes in the last, well, decade or so are thoroughly documented. So, it's nice to be able to report some good vibes. Its share price leapt nearly 15% yesterday, primarily in response to the appointment of new CEO Lip-Bu Tan.

Meanwhile, an engineering manager at Intel posted some celebratory images from the company's Arizona fab marking an "Eagle has landed" moment for the crucial 18A node, an indication the new node might already be in mass production.

I as tap these words out, Intel's share price sits at $23.88, over 20% higher than the circa $19.50 it was trading at on Tuesday. That's still a long way of the $65-plus level it hit in April 2021 and also well below the $42 it hit way back in 2000.

Long story short, Intel has a long way to go yet if we can say it's turned any sort of corner. Indeed, by the estimates of most financial experts, Intel's break up value, or how much money you'd net by breaking the company up and selling all its assets to the highest bidder still comfortably exceeds its current market cap, or its current overall valuation based on its share price.

Whenever a break up value exceeds market cap, it's an indication that the markets are pessimistic about the company's prospects. So, in net terms, that remains the case today, even after a double-digit bump in the stock price. But up is up, and Intel is definitely on the up this week.

But what of that 18A moment? Inevitably, the Linkedin post in question seems to have already been taken down. But not before it was captured by several observers.

In it, Intel engineering manager Pankaj Marria posted an image of a group of engineers in front of an 'Eagle has landed poster' and commented, "proud to be part of the Eagle Team, leading the way in bringing Intel 18A technology to life! Our team was at the forefront of running the initial lots right here in Arizona, marking a key step in advancing cutting-edge semiconductor manufacturing."

He also added, "the Eagle has landed, and this is just the beginning! Developed and Made in America the World’s smallest Node."

Several observers are taking Marria's comment of "running the initial lots" to mean that mass production had started. Frankly, it's hard to know if that's a reasonable interpretation of the Linkedin post. But at the very least, there seems to be clear positivity inside Intel about the health of 18A. And that is a very, very good thing for not only Intel itself, but the chip industry as a whole.

Currently, TSMC dominates the customer chip manufacturing market in terms of cutting-edge silicon. That's made access to TSMC's fabs critical, it's pushed up prices, and arguably lead to the current situation where it's hard to find a decent graphics card to buy, never mind how much you'll have to pay for it.

Intel is planning to take on TSMC directly with its 18A node, offering it to everyone from Nvidia to AMD as an alternative manufacturing solution. If 18A succeeds, it will relieve much of that pressure, industry wide, and might (just might!) help return the GPU market to something approaching normality.

Intel is also planning on using 18A for its next-gen Panther Lake mobile and Nova Lake desktop CPUs. No doubt if 18A really is killer, it could be used for a future Intel GPU architecture, such as Celestial or Druid.

Long story short, 18A is crucial. Go, 18A, go! Please...

"Intel 18A is now ready." So proclaims a new landing page on Intel's website for the company's all-important new 18A chip production node. But what does it mean for the PC?

We already knew that Intel's next laptop chip, Panther Lake, is due to be made, at least in part, on the new 18A node. That's supposed to be going into volume production later this year, though we're not expecting 18A-powered laptops until early 2026.

Intel's next desktop CPU, codenamed Nova Lake, is likewise a 2026 product according to Intel. So, what to make of Intel's claims of immediate 18A readiness?

The new web portal is arguably more aimed at customers for its nascent foundry business than bigging up its own chips. "Intel 18A is now ready for customer projects with the tape outs beginning in the first half of 2025," the website says.

In terms of in-house chips, Intel does call out the Clearwater Forest server CPU as an example of "18A in action" on the new website. Clearwater Forest was originally on Intel's roadmaps as a 2025 product. However, at the end of January, Intel pushed Clearwater Forest back to the first half of 2026, which doesn't exactly seem like a huge vote of confidence in 18A.

Of course, 18A is the final part of Intel's bold "five nodes in four years" plan (also known as 5N4Y), which kicked off with Intel 7, then went to Intel 4 as used in the Meteor Lake mobile CPU, followed by Intel 3, Intel 20A and finally Intel 18A.

It was mid-2021 when Intel's then-new-but-now-departed CEO Pat Gelsinger floated the idea. That means 18A needs to be ready for summer 2025 to be on track. With that in mind, the cynic might conclude that 18A's announced readiness for customers, all the while Intel doesn't seem to be able to get one of its own chips on 18A ready, is something of a PR stunt.

Arguably, you might conclude that about 5N4Y in general. Looking at each of the nodes in turn, Intel 7 is a version of the same 10nm technology that Intel had been struggling with for a decade, so not really a new node.

Intel 4 was certainly new, albeit a rebrand of the planned 7nm node. Calling Intel 3 a new node is a bit of a stretch, being as it is a revised on Intel 4. You could say the same about Intel 20A and 18A, but Intel cancelled 20A in any case.

So, at the very least, Intel is down to four nodes in four years. But taken in the round, of the five "new" nodes, only Intel 4 and Intel 18A are unambiguously new. And it remains to be seen whether any chips on 18A are actually available within Intel's self-imposed time frame.

Still, if 18A is as good as Intel is cracking it up to be, whether it's complete "ready" now or early next year probably doesn't matter. It'll be a great node that will not only enable some really competitive chips for Intel, but surely have customers queuing up to have their chips manufactured by the only real alternative to Taiwanese mega-foundry TSMC.

Among other advantages and refinements of the 18A node, Intel is making the following claims:

• Up to 15% better performance per watt and 30% better chip density vs. the Intel 3 process node.
• The earliest available sub-2nm advanced node manufactured in North America, offering a resilient supply alternative for customers.
• Industry-first PowerVia backside-power delivery technology, improving density and cell utilization by 5 to 10 percent and reducing resistive power delivery droop, resulting in up to 4 percent ISO-power performance improvement and greatly reduced inherent resistance (IR) drop vs. front-side power designs.
• RibbonFET gate-all-around (GAA) transistor technology, enabling precise control of electrical current. RibbonFET allows further miniaturization of chip components while reducing power leakage, a critical concern for increasingly-dense chips.

If 18A really does deliver on all that, it'll certainly be highly competitive with anything TSMC has to offer. Broadly, Intel 18A is thought to be less dense in terms of logic gates than TSMC's upcoming N2 node and more akin by that metric to the N3 node that TSMC has been banging out for about 18 months now.

However, another key measure is SRAM density. SRAM cells are used to provide critical on-chip cache memory. It was thought until recently that 18A was at best on par with TSMC N3 for SRAM density. However, it's recently emerged that Intel 18A in fact offers pretty much exactly the same claimed SRAM density as TSMC N2.

Meanwhile, TSMC is not planning to include backside power delivery until its own A16 node. Long story short, backside power relocates the power interconnects from the top of the chip to the underside of the silicon layer, thereby separating it from signal interconnects. This reduces interference and shortens the distance power has to travel, which improves efficiency and performance.

As above, Intel 18A has backside power, which could be a big advantage over TSMC N2. Ultimately, we'll have to wait and see. But we absolutely have our fingers and toes crossed that 18A works out. The alternative could be very bad for Intel. Gelsinger did say, after all, that he had "bet the whole company on 18A."

And the way the chip industry is going these days, "very bad for Intel" could quickly turn into even more expensive chips for the PC.

We need as much competition in the industry as we can get. So, we generally agree with the sentiments an Intel engineer posted and then removed from Linkedin earlier this week, which basically boil down to not giving up on Intel just as it is about to turn things around with 18A.

Intel won't launch a new desktop CPU until 2026. That's the clear implication from comments made this week by the company's interim co-CEO.

Speaking on an earnings call to an assembly of the usual financial analyst types, Michelle Holthaus said, "2026 is even more exciting from a client perspective as Panther Lake achieves meaningful volumes, and we introduce our next-generation client family code-named Nova Lake."

Nova Lake is, of course, Intel's next desktop architecture, the replacement for the current Arrow Lake chips including the Intel Core Ultra 9 285K. Now, we only reviewed that back in October last year. So Arrow Lake is still very new. You wouldn't normally expect a new CPU architecture anytime soon.

In that sense, the fact that Nova Lake isn't due until next year—and we don't have any clear indication in what part of 2026—is hardly a surprise. That said, there are reasons to think Intel might be keen to move things along more quickly than usual when it comes to Nova Lake.

By way of example, Intel released its new Meteor Lake laptop chips in December 2023, only to follow them up in very short order with the Lunar Lake family in September 2024. Arguably, Lunar Lake wasn't a direct replacement for Meteor Lake. But, still.

Moreover, Meteor Lake wasn't super competitive for various reasons and you could say the same of Arrow Lake. Likewise, Arrow Lake is built using predominantly TSMC silicon, and Intel is on record saying that hurts its profit margins.

So, Intel has plenty of reasons to get Nova Lake out the door ASAP. The potential catch is that Holthaus also confirmed that at least some Nova Lake CPU models will have silicon made by Intel on its new ultra-advanced 18A node.

"Nova Lake will actually have die both inside and outside [Intel Foundry] for that process [18A]," she said. As we reported earlier, Intel actually plans to release Panther Lake mobile chips on 18A later this year.

But as we also reported in the same story, Intel only made about 5% of its chips on the latest Intel 4/3 node in 2024, despite releasing the first Intel 4 product, Meteor Lake, in 2023. The point being that Intel has hardly been ramping up its new nodes at speed.

That implies limitations as to Intel's ability to crank out lots of chips on 18A and therefore its ability to bring Nova Lake forward. In the meantime, we might expect to see Arrow Lake get some kind of refresh, though there are also rumours that Arrow Lake Refresh was planned but has now been canned.

Arguably, it doesn't really matter. Arrow Lake is what it is and a refresh is highly unlikely to turn it into an AMD Ryzen killer. It's not a terrible CPU by any means. It's just a little disappointing. Whether Nova Lake will change that, whenever exactly it arrives, is anyone's guess.

Intel has been talking numbers this week with the announcement of some reasonably positive financials for the final quarter of 2024 (you can read the full transcript here). With that comes the usual chat around product and official confirmation that the Panther Lake mobile CPU is still due in the second half of 2025 using Intel's all-important new 18A silicon node.

But Intel said 18A wouldn't necessarily win all future CPU designs and also revealed how slowly the company's transition to advanced silicon has been. In 2024, just over 5% of Intel's internal chip manufacturing was on its latest EUV-based nodes.

"Looking ahead to the rest of the year, we will strengthen our client road map with the launch of Panther Lake, our lead product on Intel 18A in the second half of 2025," Intel's new interim co-CEO Michelle Holthaus said on the latest earnings calls with the usual money men and spreadsheet soothsayers.

However, Holthaus also implied that Panther Lake using Intel's new 18A node doesn't automatically mean that all of the company's future CPUs will be coming back in house. For its next-gen desktop CPU Nova Lake, it seems Intel is planning to split production between its own 18A silicon and likely a TSMC node, perhaps N2.

"We look at each generation of products based on what's the right product, what's the right process, what's the right market window and what allows our customers to win. So, for Panther Lake, that was 18A," Holthaus explains.

"Then as you look forward, to our next-generation product for client after that, Nova Lake will actually have die both inside and outside [Intel Foundry] for that process. So, you'll actually see compute tiles inside and outside," she said.

During the call, Holthaus made generally positive noises about the 18A node which is critical to Intel's future, not only for its own chips but also as a foundry service offering to customers in competition to Taiwanese megafoundry TSMC. But the fact that Intel doesn't plan to make Nova Lake exclusively on 18A is, perhaps, revealing.

It is not, however, entirely surprising. Despite plans to begin selling Panther Lake chips built on 18A later this year, by Intel's own admission it has barely begun scaling up production of chips on its latest Intel 4 and Intel 3 nodes.

"EUV wafer revenue grew from 1% of total revenue in 2023 to greater than 5% in 2024," revealed Intel's CFO David Zinsner. Intel 4 is company's first commercially available node to use EUV or extreme ultraviolet lithography and is used for the Meteor Lake family of Laptop CPUs.

However, the follow up to Meteor Lake, Lunar Lake, somewhat surprisingly reverted to TSMC for production, as did the Arrow Lake desktop family including the Intel Core Ultra 9 285K.

In other words, Intel is planning on cranking out 18A chips shortly despite having barely begun the transition to Intel 4 (or Intel 3, which is a derivation of Intel 4). Given that previous CES, Pat Gelsinger, admitted to betting the company on 18A it's needed to keep the new node on track and get product actually built on it whatever state the manufacturing is in with the Intel 4/3 nodes.

With all that in mind, it remains a very mixed picture at Intel. Even if the company does manage to release some Panther Lake chips on 18A later this year, questions will remain about the viability of its production technology.

Intel doesn't seem to have been able to ramp Intel 4 and Intel 3, despite releasing Meteor Lake well over a year ago. So, who knows what will happen with 18A.

We're still getting our heads around Intel's latest Arrow Lake desktop CPUs (actually you could say, so is Intel). But here comes news of its successor, namely Nova Lake.

It's been spotted on a shipping manifest by X user X86 is dead&back (via Tom's Hardware). The manifest provides little detail, though it seems to indicate the chip in question may be a low-end i3 variant. Of course, Intel no longer uses the "i" prefix in its CPU nomenclature, so that may be some kind coincidence or just a hangover from the previous branding era, a placeholder or similar.

Anyway, Intel itself hasn't revealed much about Nova Lake other than to say that it will mostly be built in house, with a few isolated models possibly farmed out to TSMC.

Due around the 2026 or 2027 time frame, Intel hasn't explicitly said which production node will be used by Nova Lake, but given that it says its next mobile CPU, Panther Lake, will be produced internally on 18A silicon, and that Panther Lake comes before Nova Lake, 18A at least seems like a safe bet.

While there are some rumours that Nova Lake will be built on Intel's future 14A node, that is rather speculative, what with Intel yet to prove that 18A is working well. Moreover, actual test chips like this built on 14A doesn't seem likely.

The other big question involves Nova Lake's construction. It'll be a chiplet design again, but what kind of chiplets? Arrow Lake's architecture arguably suffers due to the memory controller being located on a separate die from the CPU cores, introducing performance-sapping latency.

Intel's more successful Lunar Lake laptop chip used just two chiplets and put the CPU cores on the same die as the memory controller. It's unclear exactly how Panther Lake and Nova Lake will be segmented. But some rumours claim Panther Lake will have CPU cores and memory controller on the same die, while Nova Lake will not.

As for the specifics of those cores, it's thought Nova Lake will get Panther Lake's Panther Cove P-cores, renamed to Coyote Cove, plus Arctic Wolf E-cores. At this stage, very little is known about how those cores perform. But, arguably, it'll be that 18A silicon that defines Nova Lake as much as anything.

How good will 18A be? Will Intel actually use 18A for Nova Lake? After all, Arrow Lake was originally meant to at least in part built on Intel's 20A node before Intel summarily ditched that idea and shifted the whole thing bar the base tile to TSMC, with the CPU cores made on TSMC's N3B node.

Anyway, with Arrow Lake hardly setting the PC enthusiast world alight, we dare say Nova Lake can't come soon enough. Lunar Lake shows that Intel hasn't totally lost its mojo, so there's no reason why Nova Lake can't be a banger. But it does rather need to be just that to get Intel back on track and stop what currently feels like a bit of an inevitable slide towards an eventual demise.

A leaked roadmap of Dell's laptop plans shows the system builder is planning on supporting five new CPU architectures in new Dell machines over just three years, with the current Raptor Lake and Meteor Lake models being given a serious overhaul.

The detailed plan was discovered by Videocardz, in the form of a 300-page document that goes into great depth about Dell's plans for its XPS series of laptops, powered by Qualcomm's Snapdragon X processors. Amongst the jumble of pages is a chart that shows the projected timescale for implementing Intel's next generation of CPU architectures.

At the moment, Dell uses a combination of Raptor Lake (RPL-H, RPL-P) and Meteor Lake (MTL-H) chips for all its XPS laptops, but this year will introduce an updated XPS 13 model, sporting one of two Arm-based Snapdragon X CPUs. Later in 2024, that same model will be updated again but this time, sporting an Intel Lunar Lake chip.

The larger XPS 16 range will sport Arrow Lake (ARL-H) processors in 2025, before swapping them to Panther Lake (PTL-H) chips in 2026. XPS 14 laptops will also get Arrow Lake next year but according to the chart, that model will be superseded by two new designs: one will be Panther Lake like the XPS 16 but the other will sport a second generation of Snapdragon X processors.

Still with me? Good. Other snippets tucked away in that one chart are the expectation that Nvidia's next generation of mobile GPUs (i.e. Blackwell RTX 50-series) will appear in 2025 and that Intel's eventual successor to Panther Lake, called Nova Lake, is expected to make an appearance in 2027.

So, to summarise things a bit, a total of five different architectures in the space of just three years. That's a radically progressive strategy by Dell, even though it regularly updates its portfolio every time AMD or Intel releases a new CPU. Meteor Lake, Arrow Lake, and Panther Lake probably share a common platform or at the very least, share some common elements.

But that's still an awful lot of changes in a relatively short period and I should imagine it's going to cause no end of confusion to the casual buyer, trying to understand the differences between the various models.

It's also a little surprising that Qualcomm is planning on a revision to its Snapdragon X Elite chip when it's barely hit shelves and there's no indication as to how well it's going to be received by the laptop market. Qualcomm and Dell obviously feel confident that it's going to do well enough to warrant a model refresh, though they're perhaps been rightly cautious by waiting almost two years before doing it. 

The one caveat to all this, however, is that the leaked document is from August last year, almost 12 months old now, and so the roadmap could well have changed in the intervening time. The doc mentions an Nvidia laptop GPU refresh that clearly isn't going to happen, so not all can be taken as gospel. Still, it's an interesting insight into just how tricky it's going to get for laptop makers with just a third extra player in the Windows device market, with Qualcomm joining AMD and Intel. Just imagine what it'll be like when the other Arm-based chips arrive.

The Meteor Lake laptops we've reviewed so far have been pretty good, fine at 1080p gaming, though not especially great in terms of power consumption. Hopefully, each new Intel architecture will keep on improving game performance and make the chips a tad kinder on the battery. Just how good the first-gen Snapdragon X Elite CPUs Will be in the final reckoning, though, is still up in the air. Qualcomm is confident, and maybe Dell is, too, but it won't be long before we find out whether its silicon can really compete.