Following outstanding investigative journalism by the likes of Gamers Nexus and Hardware Unboxed, the saga of Principled Technologies’ flawed — and now mostly corrected — competitive testing of the Intel Core i9-9900K CPU has ended with a whimper. Why? I’ll cut right to the chase. When the benchmarking house agreed to test the AMD Ryzen 2700X with all of its 8 cores and 16 threads actually enabled, Intel no longer holds a claim of being “up to 50% faster” than AMD’s gaming flagship. That lead has been slashed to only 17% at best and 12% when averaged out across all games tested.
And the 9900K will still cost consumers 66% more than its Ryzen competitor. That’s before you buy an adequate cooler for the processor, as the 9900K ships without one in the box.
Let’s do a quick recap for those of you who missed the drama. Following its New York launch earlier in the week, Intel published a comprehensive set of gaming benchmarks on its Newsroom site (10 days before any embargoed reviewers could dispute the results) from a study it hired third party testing house Principled Technologies to conduct.
The benchmarks comprised 19 games and included various CPUs like Ryzen Threadripper 2950X, Ryzen 2700X, Intel Core i7-8700K and Intel’s brand new Core i9-9900K. The study allowed Intel to conclude that its $480 CPU was up to 50% faster than AMD’s $299 Ryzen 2700X. Principled Technologies outlined their testing methodology in painstaking detail, which of course allowed the tech press to quickly pick apart several crucial flaws in their testing procedures, rendering the results irrelevant.
Those issues included, but weren’t remotely limited to:
- Testing all AMD CPU systems with Ryzen Master’s “Game Mode” activated, which is only intended for Threadripper and disables half the CPU cores.
- Using AMD’s stock Wraith cooler for the Ryzen 2700X, but upgrading Intel’s 9900K to a superior Noctua CPU cooler.
- Not enabling an XMP profile for AMD systems, resulting in loose memory timings. XMP was enabled on Intel systems and memory was configured properly.
- Testing the graphics benchmark for Civilization VI and not the more CPU-reliant AI benchmark.
Even if you don’t follow the technicalities of CPU testing or enthusiast gaming, you only need to scan that first bullet point to see the glaring issue.
When confronted by enthusiast gamers and the tech reviewer community, Principled Technologies acknowledged some of these faults and pledged to run their entire benchmark suite again with a revised testing methodology. To the company’s credit, it not only did so but published the results.
Intel originally staked its “up to 50% faster” claim on Ashes of the Singularity, a demanding multi-threaded benchmark. You can imagine what happens to the Ryzen 2700X when 4 of its cores and 8 of its threads are disabled. With the chip running at full capacity, that performance advantage dropped to 17%.
The overall performance advantage Intel’s i9-9900K now enjoys following the corrected testing methods? About 12% averaged across all 19 games. And it’s important to point out that Principled Technologies still didn’t address proper memory configurations, handing the Intel platform an advantage that could further shrink when tested correctly.
Intel PR sent along yet another tone-deaf statement to press after Principled Technologies published the revised results:
“Given the feedback from the tech community, we are pleased that Principled Technologies ran additional tests. They’ve now published these results along with even more detail on the configurations used and the rationale. The results continue to show that the 9th Gen Intel® CoreTM i9-9900K is the world’s best gaming processor. We are thankful for Principled Technologies’ time and transparency throughout this process. We always appreciate feedback from the tech community and are looking forward to comprehensive third party reviews coming out on October 19.”
The thing is, no one — not even tech press at the launch event — ever disputed that the i9-9900K would be the world’s fastest gaming processor. With reviews always emphasizing a price-to-performance value proposition, the big question was how much faster? Intel of course uses the word “best” and I think that’s up for debate. Will it be the best when thermals are taken into consideration? When the 66% higher price is taken into consideration? When reviewers and consumers evaluate the added cost of a 3rd-party CPU cooler. We’ll certainly find out — from truly independent parties — on October 19th.
Let’s block ads! (Why?)