Intel "Bartlett Lake-S" Gaming CPU is Possible, More Hints Appear for a 12 P-Core SKU

[AleksandarK]

Intel's "Bartlett Lake-S" architecture, previously only offered for edge and networking deployment, may spawn a 12 P-core variant for gamers that eliminates efficiency cores entirely. This hopeful configuration would specifically target applications that benefit from consistent single-threaded performance and deterministic core behavior, addressing a market segment underserved since Intel's transition to hybrid architectures. Recent software support developments strengthen this speculation, with diagnostic utility AIDA64 explicitly adding "improved support for Intel Bartlett Lake-S CPU" in its 7.65.7404 beta release notes from April 13, 2025. This update precedes any consumer launch announcement, suggesting possible platform expansion. MSI-affiliated overclocker Toppc amplified these rumors by highlighting the AIDA64 changelog while referencing undisclosed developments under NDA, a pattern historically preceding consumer product launches.

The rumored gaming-oriented CPU would leverage the LGA 1700 platform compatibility, enabling drop-in upgrades for existing 600-series and 700-series motherboard owners. Unlike the current flagship Core 7 251E with its 8P+16E configuration, a pure performance-core implementation would eliminate the Windows scheduler complications that sometimes impact frame timing in latency-sensitive games. Current hybrid designs force game engines to navigate complex thread scheduling across heterogeneous cores, with performance-critical threads occasionally migrating to efficiency cores during intensive scenes. A homogeneous 12 P-core architecture would eliminate this behavior, providing stable thread assignment and potentially reducing the 99th percentile frame time variances that affect perceived smoothness in CPU-bound titles.



View at TechPowerUp Main Site | Source

 

[Stephen.]

Will they reintroduce HT? since now it appears to target 1851 socket.

Previous rumours were pointing at socket 1700

 

[Quicks]

Make them all performance cores, these silly efficient cores is pointless and aren't efficient at all if it raises wattage, temperatures and compatibility, for minimal gains.

Go back to the drawing board, or just stick to the old trusted way of doing things.

 

[Nanochip]

The rumor was Bartlett lake would be on lga1700 not 1851. If that changed when did it change ?

 

[rattlehead99]

Make them all performance cores, these silly efficient cores is pointless and aren't efficient at all if it raises wattage, temperatures and compatibility, for minimal gains.

Go back to the drawing board, or just stick to the old trusted way of doing things.

You have slept on Arrow Lake haven't you?
The E-Cores have 14th gen IPC/PPC and they make the CPU really cool.

 

[Hyderz]

What’s this intel giving more p-cores… unreal

 

[kondamin]

Ah a 500W idle desktop cpu, just what i was waiting for.

 

[dj-electric]

ARL-S doesn't much of a power consumption issue. It surely doesnt have much of a core-count issue. ARL-S has a per-core compute issue it has to solve.
Frankly, unless this is treated, I don't see having 12 P-cores as a game-changer. No pun intended.

 

[SL2]

Nothing here tells me 12 P and up to 6 GHz, let alone unlocked. Without both I can't imagine much improvement for games. More cores won't compensate for lower Hz, obviously.

The E-Cores have 14th gen IPC/PPC and they make the CPU really cool.

What a lovely flamebait, I think you know that's wrong lol

 

[Verpal]

Is there even a viable market for bartlett lake if it is indeed on LGA 1851 instead of previously rumored LGA 1700? The performance uplift seems too small, i can only see something happen if intel price it aggressively.

 

[john_]

Make them all performance cores, these silly efficient cores is pointless and aren't efficient at all if it raises wattage, temperatures and compatibility, for minimal gains.

Go back to the drawing board, or just stick to the old trusted way of doing things.

That Hybrid design gave them back the market. Until the 12th gen, Intel was losing the battle against AMD, because AMD was offering 16 core CPUs when Intel was having a difficulty to even bring out a 10 core CPU. And it was understandable considering AMD had a significant manufacturing advantage thanks to TSMC and Intel's manufacturing problems. With those Hybrid CPUs and those marketing cores, Intel managed to match AMD in core count and even surpass AMD in core count latter. Where AMD was offering 6 core CPUs, Intel was advertising 10 core CPUs. AMD was losing this battle, until X3D CPUs where introduced and Intel started facing it's own problems with degradation of 14K CPUs and performance regression with it's newer CPUs.

Those silly pointless efficient cores that I was calling from day one "marketing cores" was a genius move and I would have given to those who thought it, the biggest bonus Intel have ever given to an employee. And considering they dropped HyperThreading, I guess they will go the "efficient way" by adding more and more E cores in their future processors. Because almost no one reads "This CPU comes with X threads". Almost everyone reads "This CPU comes with X cores". And 99% of people can't even realize that their 10 core monster is slow, because only 2 of those cores are P cores.

Those little cores are a huge marketing advantage.

 

[dyonoctis]

And 99% of people can't even realize that their 10 core monster is slow, because only 2 of those cores are P cores.

Those little cores are a huge marketing advantage.

As those get better, this will become a non issue for the thin and light market that they target. A few of the computers at work handling the raster image processor of our printers are running on Intel 8th/10th gen, which are probably slower than a current e-core cluster. There's been very few cases were those computers felt slow

 

[Shihab]

"Gaming" CPUs should be all e-cores.
GPUs are already wasting enough kw.hrs. And me thinks we should stand back a bit and consider what we are wasting them on...

 

[ncrs]

We have an official Intel source (from the Linux kernel) confirming Bartlett Lake is Raptor Cove:

#define INTEL_RAPTORLAKE IFM(6, 0xB7) /* Raptor Cove / Enhanced Gracemont */
#define INTEL_RAPTORLAKE_P IFM(6, 0xBA)
#define INTEL_RAPTORLAKE_S IFM(6, 0xBF)
[...]
#define INTEL_BARTLETTLAKE IFM(6, 0xD7) /* Raptor Cove */

I have serious doubts about LGA 1851 compatibility. Meteor/Arrow Lake's platform differs significantly from LGA 1700.

 

[john_]

As those get better, this will become a non issue for the thin and light market that they target. A few of the computers at work handling the raster image processor of our printers are running on Intel 8th/10th gen, which are probably slower than a current e-core cluster. There's been very few cases were those computers felt slow

Intel knew from the beginning that a few P cores are more than enough for everybody. And 8 P cores are enough for gamers. So they build their hybrid system around that principles. 8 P cores for those who know what they buy, less P cores and plenty of E cores for everyone else.

And to give you a better example than your 8th and 10th gen Intel cores, until recently I was running a 6 core, 14 years old, AMD Thuban (AM3+) and it was feeling smooth and fast. And I would have kept it running if the motherboard haven't decided to say goodbye to the world. The only way to realize that it was slower in about everything typical, like browsing and opening word and excel files, next to an AM4 system, not to mention AM5, was to put those next to each other. Then you could spot the 1-2 sec of difference in opening apps for example, even when running the same type and speed storage subsystem in both AM3+ and AM4 systems. So, yes, most people will buy that 10 core monster and most of them will feel their system fast, until they decide to do something more compute heavy. But even then they will probably think that something else is slowing down their system. Maybe a virus or something. Those who will buy that 10 core munster to do something compute heavy from day one, they will get disappointed.

Now I understand what you mean by saying that E cores are becoming faster. Atom cores are becoming faster the last 15+ years. But the thing is that those cores will ALWAYS be slower than P cores, either Intel P cores or even AMD P cores. Except if somehow Intel comes out with some huge performance advances and AMD drops behind. And considering that single digit performance differences some times differentiate models, having slower cores in a CPU means you don't get the performance you think you get. You think that you are "X fast" in general, but in fact you are "X fast" in single/low threaded applications and "X-10%" or "X-20%" slower in heavy threaded applications compared to a P cores only model.

P.S. In the past there where a couple of videos showing that systems with E cores wheren't as smooth as expected, but there was a kind of stuttering observed in their performance. Maybe the switch of the CPU between P cores and E cores in handling tasks wasn't as smooth as it should be. But those videos never really gone viral, so probably they where individual cases and not something common.

 

[WatchThe80s]

12P core, 200W on full load at all core 6.0GHz... ah, sweet dreams.

 

[Fatalfury]

 

[TSiAhmat]

"Gaming" CPUs should be all e-cores.
GPUs are already wasting enough kw.hrs. And me thinks we should stand back a bit and consider what we are wasting them on...

As far as I know: e-cores aren't efficient cores (in energy) but efficient in space allocation.

 

[Upgrayedd]

E-cores are cool if the scheduler worked. But also I don't need 16 flippin E cores. Try like 4.

 

[TristanX]

Bartlett is not S1851 and not gaming, just new niche NEX products. For gaming Intel is improving Arrow Lake (APO) and create Arrow Lake Refresh, both confirmed by company leaker

 

[JustBenching]

having slower cores in a CPU means you don't get the performance you think you get. You think that you are "X fast" in general, but in fact you are "X fast" in single/low threaded applications and "X-10%" or "X-20%" slower in heavy threaded applications compared to a P cores only model.

That's just absurdly wrong. A 12 pcore chip will get blasted by an 8+16 chip while occupying similar die space. There is nothing that a full pcore chip will be faster at, the heck are you talking about?

 

[Shihab]

As far as I know: e-cores aren't efficient cores (in energy) but efficient in space allocation.

We do know that e cores consume less power than p ones. And we know that they are energy-efficient in simple workloads. It doesn't need advanced maths or any experimentation to conclude that they definitely consume less energy in workloads with fixed-duration functions (those that run for the same amount of time, regardless of performance, like games) by the sole virtue of drawing less power.
They are only energy inefficient when it comes to heavy workloads with no fixed duration, like offline simulation, offline rendering, and such.

 

[john_]

That's just absurdly wrong. A 12 pcore chip will get blasted by an 8+16 chip while occupying similar die space. There is nothing that a full pcore chip will be faster at, the heck are you talking about?

So, an application that uses for example max 4 threads, will be faster on a 2P+4E CPU than on a 4P CPU. An application that uses max 8 threads will be faster on a 4P+8E CPU than on an 8P CPU. An application that uses max 12 threads will be faster on a 8P+8E CPU than on an 12P core CPU etc.

When Intel goes to 12P core CPUs people like you will be screaming about how much better a 12P core CPU is compared to an 8P core CPU because X game/application can utilize all those extra P cores. For now you will be simply saying "That's just absurdly wrong".
I am fine with that.

 

[InVasMani]

You have to wonder how these will be clocked and what the cache looks like if Intel's positioning these as 8P + 16E against a 10P or like a 6P + 8E against a 8P of these types. That essentially means a hypothetical 9P+12E is what the 12P version would be pitted against. These must scaling higher frequency and have a larger helping of cache per core to make any sense. If they do that though they could certainly perform in scenario's where ST performance is a more dominant determination of performance as a whole versus load balancing with higher peak MT and mix of ST that's "thoughtfully" utilized.

They should probably consider something like 6P, 9E, 12LP and different cache structures on each that are balanced if they continue on with or come back to the hybrid design unless they plan to outright ditch it for all P cores again.

If Intel is going all P cores again they need to heavily modify the cache per P core if only offering that many cores especially with the removal of HT already. To me it appears like that probably what they've decided to do with these options is a more P core only design with higher cache per core, but these seem aimed pretty much strictly at people that want more the scenario I mentioned above and don't mind what they give up in place of it. That's perfectly fine there is room for both options to coexist within the market. Different strokes for different folks.

I wonder if Intel potentially consolidated 8P + 12E into a 12P design that doesn't have the software level threading contention in some way by basically merging the 16E into the existing 8P hardware design frame work in sort of a interpolated design crossover because that's entirely plausible if you interpolate the two you could end up at 12P cores that are more powerful than current P cores. It would take some careful engineering balancing, but it absolutely seem like it could've been sorted out. It's a bit like what AMD did on GPU design side when they revamped things from VEGA to the next architecture with a few key strategically re-balancing adjustments. The details of what they've done is what matters. It's likely just a node shrink coupled with more cache per P core with Intel aiming this at a target audience that wants this particular scope of performance.

 

[JustBenching]

So, an application that uses for example max 4 threads, will be faster on a 2P+4E CPU than on a 4P CPU. An application that uses max 8 threads will be faster on a 4P+8E CPU than on an 8P CPU. An application that uses max 12 threads will be faster on a 8P+8E CPU than on an 12P core CPU etc.

When Intel goes to 12P core CPUs people like you will be screaming about how much better a 12P core CPU is compared to an 8P core CPU because X game/application can utilize all those extra P cores. For now you will be simply saying "That's just absurdly wrong".
I am fine with that.

There is nothing that just uses 2 or 4 threads. But why the heck are we even taking about 2P core cpus. Even the cheapest lowest of low end desktop i3 has 4.

I've been screaming in every thread about how crap the 12p core will be compared to 8+16 for a year now. It will be slower in everything.