Qualcomm Snapdragon X Elite "X1E80100" CPU Gets Geekbenched

[T0@st]

Last October, Qualcomm introduced Snapdragon X Elite as its most powerful computing processor for PC, but the ARM-based mobile solution is still months away from launch—officially mid-2024. Company leadership has indicated that their custom Oryon CPU—for the "thin-and-light notebook market"—could be hitting retail at the same time as Microsoft's heavily rumored "Windows 12" inauguration. Several PC news outlets have picked up on a mysterious Qualcomm "ZH-WXX" platform appearing on Geekbench Browser—the February 22 entrant seems to be a prototype notebook that sports a "Snapdragon X Elite - X1E80100 - Qualcomm Oryon CPU," and 32 GB of LPDDR5x memory.

There are no next generation operating system revelations here—the system was running a 64-bit install of Windows 11 Insider Preview. Overall Geekbench 6.2.2 tallies are 2574 (single-core), and 12562 (multi-core)—positioning the 12-core Snapdragon X Elite engineering sample just above AMD's Ryzen 9 7940HS top-end mobility-focused "Phoenix" APU in terms of performance. Geekbench Browser's "CPU Information" section identifies the alleged high-end Snapdragon X Elite processor as a "ARMv8 (64-bit) Family 8 Model 1 Revision 201" part. Average clock speeds were listed as 4.01 GHz (base frequency). Cluster 1 seems to contain eight Nuvia-designed Oryon processor cores, while Cluster 2 receives the remaining four units.




According to Wccftech: "the top Qualcomm Snapdragon X Elite CPU will come in 12-core configurations with a total of 8 high-performance and 4 efficiency-optimized cores based on TSMC's 4 nm process node. The clock speeds for the chip will be set at 4.3 GHz across 1-2 cores and 3.8 GHz for all-core while adopting a large 42 MB cache." The publication highlighted an October 2023 set of results—they reckon that "X1E80100" is closer to being the final article, when compared to the older/previously benched candidate. A leaked 3DMark Wildlife Extreme benchmark—from the same time period—indicated that the Snapdragon X Elite's Adreno integrated graphics solution was already capable of besting Team Red's Radeon 780M iGPU.




Wccftech has compiled the latest results into comparison charts—look just above. Accompanying analysis stated: "Qualcomm Snapdragon X Elite 12-Core CPU was faster than both AMD & Intel chips in single-core performance despite the latter two running at much higher clock speeds of 5.2 & 5.0 GHz respectively. The Snapdragon chip can only maintain a maximum clock speed of 4 GHz across 1 or 2 cores. All benchmarks were conducted on the same Windows platform and didn't use Linux platforms although you can find Linux tests shared by Benchleaks here. When comparing the multi-core performance, the Intel Core i7-12700K Desktop CPU ends up 4% faster while the AMD Ryzen 9 7845HX CPU ends up 5% faster. This is a very impressive multi-thread performance from the Oryon CPU architecture at just 28 W and the 80 W chip can easily outperform the two competitors. Based on the performance results, we can seethat the chip sits in a good position against the other competing chips from AMD and Intel. The Intel Core Ultra 9 185H takes a lead in the multi-threaded benchmarks while the Ryzen 9 7940HS which is essentially the same spec as the 8840HS is slightly behind but that's to be expected since it's an 8-core chip versus the 12-core Snapdragon & 16 core Intel offering."

View at TechPowerUp Main Site | Source

 

[Denver]

The worst mistake is to repeatedly refer to TDP and praise the chip's efficiency without actually knowing its true power consumption...

 

[Minus Infinity]

How is a single core of Intel or AMD apu using 55W/110W.

 

[Scrizz]

How is a single core of Intel or AMD apu using 55W/110W.

Where did you get that from? Single threaded benchmark does not mean only 1 core has power/is being used in the package.

 

[johnspack]

I think that's pretty impressive. But as for real world application,.... we'll see.

 

[Denver]

Where did you get that from? Single threaded benchmark does not mean only 1 core has power/is being used in the package.

The figures there are the TDP of each model and do not represent reality(power draw)

 

[Noyand]

The figures there are the TDP of each model and do not represent reality(power draw)

I wouldn't be surprised that Qualcomm power draw doesn't stray too much from their TDP. The M2 pro 12 cores CPU part actually consume 34-36w in real life. So with the same power draw, that would just make it a bit worse than the M2 pro with a matching core configuration on a similar node. Notebookcheck actually reviews those chips with their official power limits enforced, and geekbench seems to favor ARM in MT from what I see. The 7940Hs is faster than the M2 pro in cinbench.

If Qualcomm can "only" match AMD efficiency rather than Apple, then that whole project is dead in the water. Why go through the pain of optimizing apps for ARM if you don't get any benefits over using the widely compatible x86 CPUs...

 

[Denver]

I wouldn't be surprised that Qualcomm power draw doesn't stray too much from their TDP. The M2 pro 12 cores CPU part actually consume 34-36w in real life. So with the same power draw, that would just make it a bit worse than the M2 pro with a matching core configuration on a similar node. Notebookcheck actually reviews those chips with their official power limits enforced, and geekbench seems to favor ARM in MT from what I see. The 7940Hs is faster than the M2 pro in cinbench.

If Qualcomm can "only" match AMD efficiency rather than Apple, then that whole project is dead in the water. Why go through the pain of optimizing apps for ARM if you don't get any benefits over using the widely compatible x86 CPUs...


View attachment 336584View attachment 336585

Qualccom's 7w chip consumes 30w. How can you say that about Apple if they never advertise TDP anywhere?

 

[Noyand]

How can you say that about Apple if they never advertise TDP anywhere?

I only said that reviews measured the real life power consumption of the M2 pro at 34-36w. The Internet chose an unofficial 30w TDP for it when Apple doesn't use that kind of measurement.

Apple M2 Pro and M2 Max analysis - GPU is more efficient, the CPU not always

Apple have introduced their new SoCs, the M2 Pro and M2 Max. Following the slight lack of efficiency in the M2 compared to the M1, now we will shine the spotlight onto both of the new chips in the MacBook Pro models: the M2 Pro and the M2 Max.

www.notebookcheck.net

With the entry-level M2 Pro with 10 cores, the maximum energy consumption rose to around 28 watts (M1 Pro with 8 cores: 21 watts) - the bigger 12-core model had a consumption of around 34 watts (M1 Pro with 10 cores: 27 watts) and the M2 Max had a consumption of around 34-36 watts (M1 Max: 29 watts). Without taking that next step in the production process, Apple simply cannot magic up lower energy consumption, considering the two additional cores as well as the higher clock rates.

So if Qualcomm didn't fumble, their 28w TDP might not be a Big fucking lie this time. There's a real 5nm chip out there with the same core configuration delivering similar result around ~30w