Monday, June 16th 2025

Intel "Nova Lake-S" Series: Seven SKUs, Up to 52 Cores and 150 W TDP

Rumors of Intel's "Nova Lake-S" processors are increasing, meaning that the design is nearing completion. Expected to arrive in the second half of 2026, Nova Lake-S will offer configurations ranging from mainstream quad-core models to a flagship with 52 cores. Initial information suggests that Intel will employ a tile-based design, separating LPE cores from P-Cores and E-Cores to optimize flexibility and yield. At the top of the lineup is the rumored Core Ultra 9 model, possibly designated 385K. It will combine 16 P-cores, 32 E-cores, and four LPE-cores for a total of 52 cores, as previously rumored. With a TDP of 150 W, it will be the most powerful SKU Intel prepared for this generation. Below the flagship, Intel appears to be planning a Core Ultra 7 SKU with 14 P-cores, 24 E-cores, and four LPE cores, totaling 42 cores.

The Core Ultra 5 series may include three variants: a 28-core version with eight P-cores, 16 E-cores, and four LPE-cores; a 24-core version with eight P-cores, 12 E-cores, and four LPE-cores; and an 18-core model with six P-cores, eight E-cores, and four LPE-cores. Entry-level Core Ultra 3 parts would feature either a 16-core configuration with four P-cores, eight E-cores, and four LPE-cores, or a 12-core option with four P-cores, four E-cores, and four LPE-cores, both targeting a 65 W power envelope. All desktop SKUs are expected to feature four LPE cores on a separate die, suggesting a multi-tile package similar to Meteor Lake. Power demands will range from 65 W in entry-level segments to 150 W for high-end parts. Intel is reportedly preparing a new LGA 1954 socket even as it readies an Arrow Lake-S refresh for late 2025. Intel has also reportedly designated Xe3 "Celestial" for graphics rendering and Xe4 "Druid" for media and display duties.
Intel Nova Lake-S
Sources: chi11eddog on X, via VideoCardz
Add your own comment

104 Comments on Intel "Nova Lake-S" Series: Seven SKUs, Up to 52 Cores and 150 W TDP

#51
freeagent
Dammit, I bought an AM5 board for Zen6.. but Ultra 7 or Ultra 9 should be a nice toy..
Posted on Reply
#52
dgianstefani
TPU Proofreader
freeagentDammit, I bought an AM5 board for Zen6.. but Ultra 7 or Ultra 9 should be a nice toy..
A year away! So enough time to excuse it to the wife ;)
Posted on Reply
#53
Dr_b_
is intel set to deliver more PCIe lanes though? Why not just release a 16P Core SKU for that niche market that wants that without the hybrid problems? AVX512 coming back as well or still MIA?
Posted on Reply
#54
tfp
I thought I saw some rumor about 10.2 AVX but I can't remember if that mean AVX512 or not.
Posted on Reply
#55
AnotherReader
dgianstefaniThere were three generations of CPU supporting two generations of DDR, plus more recent rebrands on LGA-1700. During this time core counts jumped massively and ST perf went up around 30%.

Zen 5 already has hybrid SKUs with normal and C CCDs.
Single threaded performance didn't increase by 30%. The 14900K is about 15% faster than the 12900K in single threaded scenarios.

Posted on Reply
#56
RogueSix
Dr_b_is intel set to deliver more PCIe lanes though? Why not just release a 16P Core SKU for that niche market that wants that without the hybrid problems? AVX512 coming back as well or still MIA?
Rumor mill is saying "up to 36 PCIe 5.0 lanes".

Minus four lanes for DMI leaves 32 lanes for everything else/actual use, i.e. it would theoretically be possible to have a graphics card at x16 and up to four x4 M.2 slots with no dreaded lane sharing.

This would be glorious if true and I hope that AMD will at least be able to match this with consumer ZEN 6 (plus new boards).
Posted on Reply
#58
TechLurker
Dr. DroCan't wait to see how AMD will respond. If this holds these CPUs will be monsters.
AMD would just split it between 16 Full Zen Cores + 32c Zen Cores + 4 NPUs for dedicated basic AI. So technically the same 52 core count; but trading the 4 low power efficiency cores for extra AI optimization.
Posted on Reply
#59
Dr. Dro
TechLurkerAMD would just split it between 16 Full Zen Cores + 32c Zen Cores + 4 NPUs for dedicated basic AI. So technically the same 52 core count; but trading the 4 low power efficiency cores for extra AI optimization.
That would be sick :rockout:
Posted on Reply
#60
dgianstefani
TPU Proofreader
AnotherReaderSingle threaded performance didn't increase by 30%. The 14900K is about 15% faster than the 12900K in single threaded scenarios.

That's true. I was remembering the 14900KS "relative performance", and games, where the improved memory controller allowed pretty big gains.
Posted on Reply
#61
Dr. Dro
What jacks up the KS's performance is its extremely high clock speeds, though. The 14900KS has particularly aggressive clocks, it's 5.9 all core and 6.2 on the 2 turbos, like 300 MHz faster than the 13th KS (5.6/6.0) - and I can barely keep my 13900KS under nominal temperatures with conventional liquid cooling.
Posted on Reply
#62
Frizz
freeagentDammit, I bought an AM5 board for Zen6.. but Ultra 7 or Ultra 9 should be a nice toy..
I'll be switching myself around this time next year, not one to upgrade every release but I'll argue I upgraded at the wrong time lol next CPUs seem to be where the true next milestone is. That said I loved my 285K and wish I never switched to a 9950X3D on a whim.
Posted on Reply
#63
Melvis
LastDudeALiveThe Core Ultra 9 285k matched the Zen 5 9950X in fully multi-threaded workloads with an 8 thread deficit (24 vs 32). So yeah, it will translate.
I see it this way, the 9950X matched the Core Ultra 9 285 in fully workloads with a 8 Core deficit (16 vs 24) which is way more impressive since a thread is only 25% of a real core at best.

I think these new Intel CPU's are DOA honestly for mainstream desktops. Software cant even work properly with P and E cores let alone another one mixed in.....and games......well pfft. If this was to go against threadripper then I can see it make more sense, but for mainstream desktop.....nope sorry, pointless! where 8 Cores can do everything you need today and tomorrow.
Posted on Reply
#64
dyonoctis
ThomasKThose are your claims only. I'd much rather wait for actual reviews, than put my bets on some random guy's comments.

Also, I'm talking about general performance here, not the very niche and few workloads that actually fully utilize Intel's P and E cores. Most users don't need or will benefit from all those extra E-cores, hence why it's more of a PR gimmick than anything else.

Lastly, please do blindly trust Intel's judgment of how many cores the average user needs, but don't forget it is the same company that gave you nothing but 4-cores on 7 generations of products.
Eh, it's not about how core many per se, but how parallel your workload is. Intel is increasing the core count, because they expect AMD to do the same. And they can't let them have free rein at the top of the "consumer" market with their 32-core CPU.

And there's also a strong argument about how many people don't need AMD to increase the core count if you look at the current reviews of 32-core TR vs consumer CPUs. But if the core count increases on the mainstream platform, apps might be developed to take advantage of that.
www.pugetsystems.com/labs/articles/amd-ryzen-threadripper-7000-content-creation-review/

Besides, that also means that the high-end core count of today will be the mid-range core count of tomorrow. You don't have to buy the Core Ultra 9. The new Core Ultra 5 seems very solid for most people.
Posted on Reply
#65
TumbleGeorge
lexluthermiesterI've heard 44 similar to the X299 CPU lineup.
Only 1954 contacts vs 2066. Not sure that is enough.
Posted on Reply
#66
Caring1
No low power version under 65W?
Intel used to make them for workstations, generally 45W chips.
Posted on Reply
#67
dismuter
Darmok N JaladStrix Halo goes up to 256GB/s
Yes, with a 256-bit bus, which is the equivalent of quad-channel. On socketed platforms, that's currently only available with Threadripper or Xeon W.
Posted on Reply
#68
Degreco
Personally, I don't need a dozen or even more real P-cores than today's flagship CPUs offer. What I want to see from Intel and AMD is an increased boost in single thread performance in the coming architectures.

A higher core count is great if you are an encoder or crunching or running benchmarks. But my rig (and I assume most profiles too) is not at 100% core load all the time. Even when gaming or working with productivity apps, I feel 'perceived performance' like snappiness in apps and in the OS comes solely from brute single core performance (when using a modern CPU with a high enough core count, of course).

So giving a normal mainstream user (normal productivity, gaming, IMHO 90th percentile of all user profiles) a better performance than today's modern CPUs must come from single thread improvements.

That's what Apple understood years ago.

Just look at the M4 Max and the architecture. A small package but so much horsepower in SC, it's a great design. I hope AMD and Intel don't forget that, because statistically, for most users higher core counts won't dramatically change the contemporary user experience on AM5 or LGA1700/1851 that much (unless, of course and as said above, you crunch for a life). That's all IMHO, and your mileage may vary, depending on what you do with your computer.
Posted on Reply
#69
AnotherReader
dgianstefaniThat's true. I was remembering the 14900KS "relative performance", and games, where the improved memory controller allowed pretty big gains.
The KS doesn't move the needle much compared to the 14900K either. The memory controller is improved, but the 12900K can use DDR5 6400 so I suspect the gap is unlikely to be 30% without handicapping the 12900K.
Posted on Reply
#70
dgianstefani
TPU Proofreader
AnotherReaderThe KS doesn't move the needle much compared to the 14900K either. The memory controller is improved, but the 12900K can use DDR5 6400 so I suspect the gap is unlikely to be 30% without handicapping the 12900K.
Raptor Lake can do 8000 MT, sometimes a little higher.
Posted on Reply
#71
Darmok N Jalad
DegrecoPersonally, I don't need a dozen or even more real P-cores than today's flagship CPUs offer. What I want to see from Intel and AMD is an increased boost in single thread performance in the coming architectures.

A higher core count is great if you are an encoder or crunching or running benchmarks. But my rig (and I assume most profiles too) is not at 100% core load all the time. Even when gaming or working with productivity apps, I feel 'perceived performance' like snappiness in apps and in the OS comes solely from brute single core performance (when using a modern CPU with a high enough core count, of course).

So giving a normal mainstream user (normal productivity, gaming, IMHO 90th percentile of all user profiles) a better performance than today's modern CPUs must come from single thread improvements.

That's what Apple understood years ago.

Just look at the M4 Max and the architecture. A small package but so much horsepower in SC, it's a great design. I hope AMD and Intel don't forget that, because statistically, for most users higher core counts won't dramatically change the contemporary user experience on AM5 or LGA1700/1851 that much (unless, of course and as said above, you crunch for a life). That's all IMHO, and your mileage may vary, depending on what you do with your computer.
Also, note that in Apple’s larger designs, they don’t add more efficiency cores, but rather they add more performance cores. Intel does the opposite. Apple tops out at 6 E cores on M4, and actually offer fewer E cores (4) on M4 Pro and Max.
Posted on Reply
#72
Degreco
Darmok N JaladAlso, note that in Apple’s larger designs, they don’t add more efficiency cores, but rather they add more performance cores. Intel does the opposite. Apple tops out at 6 E cores on M4, and actually offer fewer E cores (4) on M4 Pro and Max.
Yep, Apple did their homework. Enhanced single core performance, multi core good enough, and all this with great efficiency.

I want that for x86 :)

Posted on Reply
#73
AnotherReader
dgianstefaniRaptor Lake can do 8000 MT, sometimes a little higher.
What's the limit for Alder Lake?
Posted on Reply
#74
matar
Great now that is an UPGREADE :clap: i am buying the 385k 52 cores and with an upgrade to my gpu ( NO not a 5090 ) the best OC edition 4090 because of my fav games use PhysX
Posted on Reply
#75
dgianstefani
TPU Proofreader
AnotherReaderWhat's the limit for Alder Lake?
About 7000, more likely 6800.
Posted on Reply
Add your own comment
Nov 13th, 2026 23:24 CST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts