Friday, December 20th 2024
Intel Arrow Lake 0x114 Microcode Already Out, No Significant Gains—We Tested
Motherboard vendor ASUS began rolling out UEFI firmware (BIOS) updates to its Intel Z890 motherboards that contain Intel's 0x114 Microcode update for Core Ultra "Arrow Lake-S" desktop processors. The new microcode was touted as bringing in performance gains to gaming workloads across the board, with Intel saying that depending on the configuration, one could expect a "roughly 3-8% performance gain." The company said that motherboard vendors should begin releasing BIOS updates with 0x114 "starting January 2025," however, it seems like ASUS is ready with public "stable" (non-beta) BIOS updates with it. We use a ROG Maximus Z890 Hero in our "Arrow Lake-S" reviews, and so promptly grabbed the version 1203 BIOS from the ASUS website, and put it to the test. This also updates Intel ME (management engine) to v19.0.0.1827.
We added our performance testing numbers to our article from yesterday (December 19, 2024), where we had tested the Core Ultra 9 285K with the latest OS-level patches for Windows 11 24H2. Long story short, we do not notice any notable performance gains with the 0x114 microcode update. 0x114 was touted as providing users with additional performance gains after all the OS- and BIOS configuration related issues had been fixed. In its pre-brief from earlier this week, Intel said that the 0x114 microcode update represented additional performance gain opportunities that the company had discovered in the process of identifying and fixing the reasons why the processors fell significantly behind Intel's performance guidance in their launch reviews in October.
We recommend you to once again read our performance testing article from yesterday, we have updated the performance graphs with 0x114 microcode update numbers, and are in the process of providing additional commentary in the article. Here's a teaser:
We added our performance testing numbers to our article from yesterday (December 19, 2024), where we had tested the Core Ultra 9 285K with the latest OS-level patches for Windows 11 24H2. Long story short, we do not notice any notable performance gains with the 0x114 microcode update. 0x114 was touted as providing users with additional performance gains after all the OS- and BIOS configuration related issues had been fixed. In its pre-brief from earlier this week, Intel said that the 0x114 microcode update represented additional performance gain opportunities that the company had discovered in the process of identifying and fixing the reasons why the processors fell significantly behind Intel's performance guidance in their launch reviews in October.
We recommend you to once again read our performance testing article from yesterday, we have updated the performance graphs with 0x114 microcode update numbers, and are in the process of providing additional commentary in the article. Here's a teaser:
23 Comments on Intel Arrow Lake 0x114 Microcode Already Out, No Significant Gains—We Tested
rog-forum.asus.com/t5/intel-800-series/arrowlake-resources/td-p/1051570/page/14
"Intel CSME Firmware Kit 19.0.0.1854v2.2 (or newer)."
Also note the v2.2 at end.
So the versions seem to make only marginal to zero differences.
I wonder how the motherboard manufacturers view this as they are suffering.
@btarunr any difference regarding consumption? Thanks.
They're identical products, it's just that Intel felt comfortable enough with the process node to officially certify higher clocks, which turned out to be a terrible idea anyway.
If you really want to be generous, deduct 1.4% from the 14900K's score to give you the 13900K's score - but that really is the best-case scenario for the 14900K in an artificially-contrived 4090@720p resolution, done that way solely to rule out the GPU bottleneck that is practically guaranteed to be your real limitation.
At a sane resolution, the scores are the same, and if there's any difference between the two chips it's more likely down to the BIOS and BIOS settings of the motherboard in question, since power limits are such a huge issue for Intel 13th/14th gen.
I still am running a Core i9 9900 KF which too this day still has some of the highest IPC values for a single thread on the market. Perfect example of Intel at it's best.
Same for buying prebuilts. Some companies will take some time to switch, even if they are faster than less technical users. If the price to performance is ok, there is a 11400F to be sold. Doesn't matter if enthusiasts don't care. 12400F, 12700F, that is also listed right now.
Heck, when Prescott was around, there were metric tons of them solds that were blowing up motherboards in a very short time. They still kept selling, then Pentium Ds and eventually they got their act together later.
So, when you have something like Arrow Lake, it is still quite a good CPU. In most situations it is very good but for gaming it is not so. Well, if you consider software needs to keep up - like what just happened with Cyberpunk patches - there is good potential there.
Arrow Lake is very good for multitasking and it keeps scaling if you give it more power - that is a very good place for Intel to be.
On the other hand, people need to understand that when the Ryzen 1600/2600 was out it was not a good gaming CPU. The 7700K was a 4core part that was substantially faster and more expensive (70% more). After two-three years - not more - software started adapting and you would have spent more money to get less performance on the newer titles. (this you can still verify by reading old and later ones)
Even if enthusiasts don't like ARL, programmers now have a competitive CPU with AMD, that makes development faster for them. So when they start optimizing tomorrow's games, they will be running on ARL. Games will run faster on ARL, possibly being slower on 12-14th than they would otherwise be if these chips had similar characteristics.