If you’re a regular reader of our hardware news and reviews, you’ll know that Intel’s much-hyped Core Ultra 200S series of processors not only failed to impress upon launch, but it seems to have been left to flag against the competition. In a new blog post today, which we’ve had early access to, Intel says that while the former is undoubtedly true, it’s been busy identifying what went wrong, with four key issues noted and fixes published.
The gaming performance of the Arrow Lakes was known to be lower than the previous generation Raptor Lakes before testing even began as Intel openly stated that this would be the case when it launched its new CPU architecture in October. The primary goal of multi-tile processors was lower power consumption in games, while still offering generational performance gains in multi-threaded workloads.
The latter was certainly achieved but once we all got the hang of it Core Ultra 9 285K and Core Ultra 5 245K (compensation for Core i9 14900K and Core i5 14600K), it was clear that Intel experienced very different gaming performance than reviewers. But after a while, and after I talked to Intel about my results, the truth of the day came out: Arrow Lake was not as good as expected and something was obviously wrong.
In the update by Robert Hallock, vice president and general manager of AI and technical marketing, he notes that Intel engineers have identified four specific issues that caused significant problems with game performance.
The PPM (Performance & Power Management) package is missing
Modern processors can operate in all kinds of performance states, but they require active support from the operating system to function properly. This is done via a PPM package – think of it as a driver that controls clock speeds and other timings, depending on the power settings the operating system is using.
Intel did not plan for PPM for Arrow Lake to appear in a Windows update, in time for reviews and resale of the Core Ultra 200S chip. The results? In Intel’s words, “Unusual CPU scheduling behavior; artificial performance increases when cores are disabled or affinitized; high run-to-run variability in benchmarks; reduced single-threaded performance; intermittent DRAM latency spikes; unexpected performance differences between Windows 11 23H2 and 24H24.”
All these issues from a “simple” power profile just goes to show how complex today’s processors really are, but the fact that Intel didn’t schedule the profile to appear in a Windows update in time for launch is somewhat annoying to read.
At least all these should be a thing of the past now, since Windows 11 update KB5044384 apparently the profile contains and Intel says it alone can handle up to 30% of the performance loss against expectations.
Intel Application Performance Optimizer failed to take effect
Intel’s processors have used a hybrid architecture (aka full performance, high-power P-cores and lower-performance, lower-power E-cores) in four generations of processors, including Arrow Lake, and they all require a custom thread scheduler to ensure that correct cores manage a game’s threads, to ensure the best performance.
Application performance optimizer (APO) is a tool that handles that process for specific games and Intel used it for their pre-launch performance data. However, the missing PPM meant that APO would simply do nothing, leaving up to 14% of game speed off the table.
With PPM now in the KB5044384 update it should work but it’s worth noting that few reviewers, myself included, ever use APO when testing an Intel processor for the first time. Oddly enough when I tried APO on a 285Kit worked so i can only assume i had the windows update then.
For actual gaming, however, it makes sense to install APO and let it optimize thread scheduling in the supported games. You get a free performance boost after all.
BSODs when starting games with Easy Anti-Cheat service
I didn’t experience this issue when reviewing the Core Ultra 9 285K and Ultra 5 245K but that’s simply because no games in our CPU benchmark suite use Epic’s Easy anti-cheat service. However, it turns out that there is a conflict between Windows 11 24H2 and older versions of EAC, which Epic has now fixed and is pushing the update out to developers.
That’s good news for fans of Fortnite, for example, but this doesn’t seem to be something symptomatic of Arrow Lake, just Windows. It may very well be that is but only with the 24H2 update, but for anyone using 23H2, the EAC fix will not change any game’s actual performance.
Performance settings misconfigured on reviewer/early BIOS
Testing Arrow Lake was an exercise in frustration, as motherboard vendors kept releasing new BIOS updates during the preview period. I noticed several differences between how motherboards were configured to run Core Ultra 200S chips – for example, the ring bus clock is supposed to be constant but one manufacturer has set it to scale down, which is required with Raptor Lake.
Intel says it had noticed that so-called “very important settings” were applied inconsistently, such as resizable BAR, Intel APO, clock calculation for ringer, memory controller ratio (aka gear), and persistent and transient power limits.
It’s not a small list and together they account for issues like aberrantly high memory latency (up to twice expected values), erratic ring clocks, high run-to-run variance for dynamic or unpredictable workloads, and no performance gain from BAR or APO.
Apparently this has also been resolved and “current BIOS releases for Intel Z890 motherboards have harmonized these settings.” However, I am currently using an MSI MEG Z780 Hero motherboard with its latest BIOS, but it still makes the ringer vary. Whether this is an Intel or MSI issue I can’t say at this point as I only have MSI boards on hand.
More to come
Intel rounds out its blog by saying that it has identified additional BIOS improvements that it is currently validating and is targeting a firmware release, with all of these updates and microcode 0x114, sometime in late January. It plans to present all these results in detail at CES 2025.
After all this is done, I will retest the Core Ultra 9 285K’s gaming performance, just to see how much better it is after all this work. If that’s enough for me to be able to recommend an Arrow Lake chip over the likes of Ryzen 7 9800X3D for games or Ryzen 9 9950X because creating content is a completely different matter.
I’m really glad that Intel has worked on solving Arrow Lake’s problems but time will tell if it’s too little, too late.
