Intel Core Ultra 7 258V beats AMD Z1 Extreme in DX12 gaming benchmark [Updated]
Lunar Lake comes out on top.
Update (September 14): Corrected erroneous analysis.
The Intel Core Ultra 7 258V (via Tomasz Gawroński) has just been tested on the Ashes of the Singularity benchmark. The results favor Intel’s Lunar Lake chip, which easily outperformed the Ryzen Z1 Extreme.
The Core Ultra 7 258V is one of Intel’s latest mobile processors based on the Lunar Lake platform. The processor features an octa-core design without Hyper-Threading. It has four P-cores and four E-cores. Meanwhile, AMD’s Ryzen Z1 Extreme is designed for handheld consoles and features eight Zen 4 cores with SMT and a very capable RDNA 3 iGPU.
According to Gawroński’s results, the Core Ultra 7 258V was 1.5X to 2X faster than the Ryzen Z1 Extreme. However, it’s important to note that the Lunar Lake chip was paired with 32GB of memory, whereas the Ryzen Z1 Extreme had 12GB. It’s hard to make a faithful comparison since we’re looking at chips designed for different segments. The Lunar Lake chip likely hailed from a laptop, whereas the Ryzen Z1 Extreme was like residing inside a handheld gaming device. The testing environment was different.
Nevertheless, we need more than just these frame rate numbers from a single DirectX 12 benchmark to know which chip performs better and more accurately. After all, several factors, like the power profile, cooling performance, battery levels, and more, could affect the chip’s actual performance, especially in mobile and handheld devices. That’s why you should wait for the results of trusted reviewers, like Tom’s Hardware, who have put these devices through the wringer more scientifically—that way, you know that the numbers you get are as objective as possible.
AMD’s Z1 Extreme is currently the only chip tailored toward gaming handhelds. It’s inside devices like the ROG Ally X and Legion Go. Meanwhile, newer devices, such as the MSI Claw 8 AI+, have already jumped on the Lunar Lake train, and we expect more competition to do the same. However, AMD has confirmed that the Z2 Extreme will be available in early 2025, so the next big fight involving Lunar Lake versus Z2 Extreme will be very interesting.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
Jowi Morales is a tech enthusiast with years of experience working in the industry. He’s been writing with several tech publications since 2021, where he’s been interested in tech hardware and consumer electronics.
-
JamesJones44 Too many * in here for this to be a comparison to draw any conclusions from. A test really needs to be done with the same RAM and video resolution at a minimum. There is roughly a 2.5x pixel difference in the tested resolutions, RAM alone is not going to make up for that difference. Both I believe have configurable TDPs as well, without knowing that, which is "better" is almost impossible to say be simply looking at benchmark numbers alone.Reply -
paviko The 2.5x more pixel to render for Intel Core is overkill, having more RAM doesn't mean it will run that much faster with higher resolution, Actually RAM doesn't affect speed for this slow mobile GPUReply -
thestryker While the Intel chip is likely carrying a heavier burden due to the higher resolution (2880x1800 vs. 1920x1080), it does have double the RAM to help it out.
This is one of the most astounding things I've read given this is a tech site. A 150% increase in the number of pixels to render isn't going to be offset by having more DRAM. The only thing memory has to do with graphical rendering is whether or not you exceed the allocation and end up with terrible performance.
BattlemageGururu said:Are the 258V systems using alchemist integrated graphics? -
Jagar123 Yeah, this article is really pushing bad data. Too many variables that don't match to come to any conclusions here.Reply -
das_stig "That’s why you should wait for the results of trusted reviewers, like Tom’s Hardware"Reply
-
SiliconFly This is probably the most stupidest tech article I've come across in recent times. A Tom's Hardware reviewer who's comparing GPUs doesn't even know he has to test it with same/similar resolutions, otherwise the GPU will grossly under perform.Reply
He's comparing Intel GPU with a resolution of 2880x1800 vs AMD GPU with a resolution of 1920x1080. That forcing the Intel GPU to process 2.5X more pixels! It's hilarious.
And the worst part is, he believes extra RAM can compensate this when he says "While the Intel chip is likely carrying a heavier burden due to the higher resolution (2880x1800 vs. 1920x1080), it does have double the RAM to help it out". It's mind numbing!
What is happening to Tom's Hardware? -
OneMoreUser The only thing this article really does is to illustrate what TH is. Most articles here are nothing but click bait, trying desperately to find ways to make interesting headlines going as far comparing apples and oranges while pretending there is a winner.Reply
Come TH - do better. -
SiliconFly I don't think any respected tech sites are dumb enough to dilute their reputation with these kinda of brainless click baits. Hurts their bottom line in the long run.Reply
Looks more like an oversight that they've allowed one of their intern/new/junior reviewer to post this foolish article without checking it. -
guraldunix As most of the comments mentioned, there's a huge pixel difference between 2880x1800 and 1920x1080.Reply
1920x1080 is about 2.1 million rendered pixels on screen and 2880x1800 is about 5.2 million rendered pixels (both numbers rounded to the nearest 100,000).
That's a huge difference in the number of pixels the GPU is rendering at any given time. How could it possibly be useful to compare two products at such different resolutions?