Friday, September 20th 2024
Intel "Arrow Lake-HX" Leaks: Up to 24 Cores and 5.5 GHz in Laptop Form
While the upcoming desktop versions of Intel's "Arrow Lake" processors are stealing all the hype, we are getting a leak about the high-performance "Arrow Lake-HX" chips for laptop space. Thanks to Jaykihn on X, we learn that Intel has prepared six SKUs for laptop enthusiasts to power workstations and gaming. The upcoming Core Ultra 200 series SKUs are 285HX, 275HX, 265HX, 255HX, 245HX, and 235HX. Starting with the lower-end Core Ultra Core Ultra 5 245HX / 235HX, we get six P-cores running at 3.1 and 2.9 GHz base speeds, respectively. For both SKUs, E-cores are clocked at 2.6 GHz at the base. Boosing the P-cores yields a maximum of 5.1 GHz, while E-cores top out at 4.5 GHz. For graphics, the 48 EUs inside the iGPU are clocked at 1.8 GHz on the 235 HX, while 245 HX runs at 1.9 GHz. THese models don't support Thermal Velocity Boost and Turbo Boost Max 3.0, while of the remaining SKUs support both features.
Moving to the middle of the stack, there are two SKUs: Core Ultra 7 265HX / 255HX. Both feature eight P-cores and twelve E-cores. The differentiator here is the clock speed. P-cores on the 255HX run at 2.4 GHz base and 5.2 GHz boost, while the P-cores on 265HX run at 2.6 GHz base and 5.3 GHz boost. E-cores have a wider gap with the 255HX running at 1.8 GHz base and 4.5 GHz boost, while 265HX has E-cores pinned at 2.3 GHz base and 4.6 GHz boost. Both of these SKUs have iGPU with 64 EUs set at 1.9 GHz.
For the highest-end part, we have Core Ultra 9 275HX and 285HX CPUs. These SKUs come with eight P-cores and sixteen E-cores. The "weaker" 275HX CPU has P-cores clocked at 2.7 GHz base and 5.4 GHz boost, while the flagship 285HX SKU has P-cores at 2.8 GHz base and 5.5 GHz boost. E-cores run at 2.1 GHz at the base, with a 4.6 GHz boost on both SKUs. The graphics output capability is similar, with both SKUs rocking 64 EUs at 1.9 GH and 2.0 GHz for 275HX and 285HX.
Source:
@Jaykihn on X
Moving to the middle of the stack, there are two SKUs: Core Ultra 7 265HX / 255HX. Both feature eight P-cores and twelve E-cores. The differentiator here is the clock speed. P-cores on the 255HX run at 2.4 GHz base and 5.2 GHz boost, while the P-cores on 265HX run at 2.6 GHz base and 5.3 GHz boost. E-cores have a wider gap with the 255HX running at 1.8 GHz base and 4.5 GHz boost, while 265HX has E-cores pinned at 2.3 GHz base and 4.6 GHz boost. Both of these SKUs have iGPU with 64 EUs set at 1.9 GHz.
23 Comments on Intel "Arrow Lake-HX" Leaks: Up to 24 Cores and 5.5 GHz in Laptop Form
Yea, Yea I know it's only on the P cores, but just sayin :)
Lets just say I'll believe it when I see it....
www.intel.com/content/www/us/en/products/docs/processors/core/core-14th-gen-mobile-brief.html
I'm more interested in low loads power consumption. Alder Lake P and H were a disaster at those.
my 10700f has a 4800 single core boost, and I don't think I saw 4800 more than for 3 seconds combined in 2.5 years.
Put another way, the new CPU line is not affected by the old manufacturing problem.
Intel just admitted problems in late July 2024, explained that oxidation problems were detected and resolved in late 2022 (apparently without any public announcement, nothing to see here).
Regarding the other issue, 'UEFI Bios voltage settings": "We have determined that elevated operating voltage is causing instability issues in some 13th/14th Gen desktop processors. Our analysis of returned processors confirms that the elevated operating voltage is stemming from a microcode algorithm resulting in incorrect voltage requests to the processor."
Since most processors took months or even years to develop faults that resulted in frequent crashes or inoperative CPUs I think we can agree it is way too soon to establish if Intel's fix completely and once and for all times resolves this issue. If CPUs degrade even with the IUEFI fix, but slower, it could take additional months to show up. And even with these faults Intel took a very long time even admitting them - so even if they just slowed down the degradations they are probably safe from people discovering that their CPU woes aren't just theirs alone. I mean, it took angry YouTube reviewers that were denied RMAs to raise the public awareness, and then the flood of reports started -:and Intel apologists are still in denial about reports on how Intel CPUs just died in render farms, other server uses...
And if underlying issue isn't just "faulty microcode requesting too high voltages", but is connected to the whole "pushing CPUs to the limit", where we actually get pre-overclocked CPUs, what are the odds this isn't the part and parcel of the new CPUs that were designed way before these issues showed up?
The overvolting was solved with clear definitions to makers of what is and is not acceptable voltage ranges which then resulted in firmware/BIOS updates to correct the issue. Some CPUs will continue to operate fine with those adjustment applied. Some are damaged beyond which adjustments can resolve the problem.
The chemistry problem was solved with an alteration to the formulation & lithography progress to resolve the oxidation issue and two other associated problems.
These problems involved only CPU's running on boards with power profiles set above Intel specs AND were manufactured during the time Intel was using that specific formulation. All currently available 14th gen CPUs are fine and will operate as expected. All future generations of CPU are not and will not be affected by the previous problems.
You can choose to be paranoid or you can use sense.
Now before anyone calls me a fanboy, AMD had a similar, though much less severe, problem with voltage related stability a few years ago with Ryzen CPUs. A similar measure was employed to resolve it and all was well. There were people being worrisome then as well. I said the same thing then. The problem was resolved and is no longer an issue. And all is well.
now that e cores suposedly has a serious make over the chip that is going to replace the N100 should be a really interesting chip
the moment the first arrow lake cpu dies even if it’s in the hands of an overclocker that delidded it with a hammer and chisel the usual big youtube pundits will be screaming murder and blame intel for it.
intel will do everything in their power to not let that happen.
Intel's financial problems are completely separate from this and in my opinion stem mostly from long term trailing behind AMD in server space. And balancing budget while relying on outside CPU fabrication for the first time in 56 years!
if there is a lot of negativity around a company even in a smaller space like tech enthusiasts that stuff will get picked up.
but sure intel overextending them selves building all the fabs is the real reason they are in trouble