Intel claims Core Ultra 200 patches improve gaming performance by up to 26% (Update: they don't)

zohaibahd

Posts: 935   +19
Staff
In context: Intel Core Ultra 200 desktop processors failed to impress some reviewers (including us) when they launched last October as the company's new flagship Arrow Lake-S lineup. While they performed adequately in compute tests, gaming was a mixed bag, with the chips being outperformed in multiple titles by the previous Raptor Lake generation.

Fixes do nothing – Update (Jan 18): We have yet to test Intel's claimed fixes for its Arrow Lake chips, however Tom's Hardware was quick to run a few gaming benchmarks on the flagship Core Ultra 9 285K. In short, the updates fail to address the underwhelming gaming performance for such a high-end chip.

In some cases, performance even decreased after the updates, while previous-gen Raptor Lake Refresh chips benefited more from the required OS update, further widening the gap. Ultimately, we can't justify how Intel is trying to spin this story beyond what can be proven and making false promises does more harm than good.

Intel quickly investigated the situation, and by December, it had a good idea of what was causing the performance hiccups. It turned out to be a combination of power plan problems, anti-cheat settings, and other random issues – though the biggest was a missing Performance & Power Management (PPM) package.

The first series of fixes landed before the holidays in a mandatory Windows 11 update alongside some publisher patches. But Intel said we'd have to wait until CES in January for the second wave – a crucial BIOS update to really unlock the Ultra 200's untapped potential. Well, that update just dropped, and according to Intel's internal tests, it's put the performance woes to bed in a big way.

There are some sizable gaming gains in Intel's numbers. For instance, in Cyberpunk: 2077, Intel claims the fixes could net a staggering 26% increase in frames per second on patched Ultra 200 systems, as reported by PC World. Meanwhile, Far Cry 6 could see up to 16% more FPS.

However, the performance hikes don't stop at games. Intel also ran the popular 3DMark TimeSpy synthetic benchmark and said patched Ultra 200 systems could see their overall scores skyrocket by up to 97%.

We'll have to run some tests ourselves before we can corroborate Intel's lofty numbers. But all the software and BIOS fixes are available now, so if you've got an underperforming Arrow Lake build, it's time to load up those updates. Just make sure you're running the latest Windows version and check in with your system or motherboard manufacturer for that all-important BIOS update.

Now that the desktop Ultra 200 has (hopefully) been dealt with, Intel can turn its attention to ensuring a smooth, high-performance release for its new Core H and HX laptop CPUs aimed at gamers and creators. Those chips are expected to start shipping soon, so we hope Intel can ensure a smooth release this time.

Masthead credit: Madness727

Permalink to story:

 
When I see "up to X%" or "Prices from ...", I'm sure what follows is a scam, possibly with some exceptions for 1% of the cases.

WTF 'up to 26%' means? What's the performance increase everyone is guaranteed to see? For some mysterious reason I'm never among those in the vicinity of the 'up to' number.
 
Great, so now companies are releasing CPUs just like game developers release games?... "Just push them out the door and we'll fix them later with a patch or two."
Even if they end up with acceptable performance, this is quite pathetic, and a dangerous trend.
Agreed. IMO, they are only doing it this way because it's much cheaper than designing and creating new silicon. "Fast, cheap, and Easy" as my wife says...
 
To be honest, I am really getting confused which fix is which now. The last time (which was not so long ago) where they said so and so fix in OS and microcode are out, but some testers did not notice meaningful improvement. So now, are they saying there is another fix? I seriously think they should consider fixing the problems TOGETHER, test it THOROUGHLY first, before sending very confusing messages out. It was already very perplexing when they said they tested so and so, and were surprised that the performance was subpar when compared to internal testings. It begs the question what kind of internal testings they were performing before, where the software and conditions are so different and yields very different results.
 
When I see "up to X%" or "Prices from ...", I'm sure what follows is a scam, possibly with some exceptions for 1% of the cases.

WTF 'up to 26%' means? What's the performance increase everyone is guaranteed to see? For some mysterious reason I'm never among those in the vicinity of the 'up to' number.
This is common practice to be honest, and typical marketing to paint the best case scenario (putting a too good to be true number in bold) while protecting themselves from lawsuits for false marketing. Since they cannot guarantee that every user will experience the same performance bump for whatever reasons, hence, the vague performance numbers.
 
Agreed. IMO, they are only doing it this way because it's much cheaper than designing and creating new silicon. "Fast, cheap, and Easy" as my wife says...
That sounds exactly like what I want from a CPU. You can keep the slow, expensive and difficult ones.
 
When I see "up to X%" or "Prices from ...", I'm sure what follows is a scam, possibly with some exceptions for 1% of the cases.

WTF 'up to 26%' means? What's the performance increase everyone is guaranteed to see? For some mysterious reason I'm never among those in the vicinity of the 'up to' number.
I saw up to 38% improvement (meaning certain outlier titles had slightly less) but I also tweaked the internship timings slightly. These new chips are amazing and will only get better as the tech matures. Intel is on to something great with this. About time. Asking with the new GPUs they are doing well.
 
Fixes do nothing – Update (Jan 18): We have yet to test Intel's claimed fixes for its Arrow Lake chips, however Tom's Hardware was quick to run a few gaming benchmarks on the flagship Core Ultra 9 285K. In short, the updates fail to address the underwhelming gaming performance for such a high-end chip.

Pretty much what everyone expected. A single obscure metric gets the 26% improvement but everything else is up 3% and down 3% so on average there's no change. Hopefully over time things slowly improve for Arrow Lake as competition is good.
 
AMD and Nvidia are headed by DEI and full of DEI hires.
That is the most ignorant post I’ve seen so far this year.
Jensen Huang is the co-founder of Nvidia. He also happens to be an electrical engineer .
Lisa Su is also an eletrical engineer, with a degree from MIT, and she’s been manager of several huge companies before taking over AMD in 2014.

You see two Asian faces and claim DEI, you’re in the wrong forum - probably room for you over at Truth Social, bring your tinfoil hat with you
 
That is the most ignorant post I’ve seen so far this year.
Jensen Huang is the co-founder of Nvidia. He also happens to be an electrical engineer .
Lisa Su is also an eletrical engineer, with a degree from MIT, and she’s been manager of several huge companies before taking over AMD in 2014.

You see two Asian faces and claim DEI, you’re in the wrong forum - probably room for you over at Truth Social, bring your tinfoil hat with you
You need to read the whole thread. My reply was sarcastic response to the clown think companies failed because of a DEI hired. It's calling out successful companies led by people of color or women of color.
 
Back