Performance and Testing Methodology
For general performance testing, we use virtually the same software we use in our motherboard testing and evaluation. Each test is run multiple times to ensure accuracy. The middle result is used in each case. The following system configurations were used for all benchmarking and testing runs.
We want to thank ASUS for providing us its ASUS Prime X299 Deluxe II motherboard to pair with our CPU today so we can give the CPU the best chance possible. Look forward to a review of the motherboard in the future.
Windows Setup
Due to the scheduler improvements to Windows 10 build 1909, we decided to use build 1909 for all system testing. While many of the applications here have been used before in our motherboard reviews, even with the same configurations, the numbers are not comparable. Those were all done on builds 1803 and 1809 using the drivers that were current then. All systems were freshly formatted, and all the latest drivers and OS patches were used.
All of the systems were updated to their latest BIOS revisions. Finally, for the Intel system, I did install the CPU microcode updates relevant to that CPU. It’s important to note that build 1903 does contain improved mitigations for several security flaws on Intel processors. However, I did not go out of my way to download any additional or optional mitigation patches. Hyperthreading (or SMT) also remained enabled for all testing.
Finally, all systems were run at stock and overclocked values. The “stock” settings are their automatic or default base/boost clocks. The boost clocks are shown in the graphs for reference. An overclocked value for each CPU was used as well, running all cores at a fixed speed shown in each table.
Comparisons
Ordinarily, I would have cross-compared the Ryzen 9 3950X, Threadripper 3960X and or 3970X with the Core i9 10980XE in reviews for each or at least, for each product family. However, AMD did not answer our request for samples of any of these processors prior to the embargo lift on these CPUs. The launch dates being the same for all the processors in question would be advantageous as we wouldn’t have to keep CPU results under wraps.
Being that TheFPSReview is a relatively new site, we do not have a massive amount of CPU’s in inventory available for testing. We’ve purchased all the mainstream CPU’s we couldn’t get samples for. However, mainstream processors are one thing. The world of HEDT CPU’s is entirely different and the costs are substantial. At the time I started working on this article, many of the retailers were still selling the Core i9 9980XE for close to $2,000 which is simply too much out of pocket.
Lastly, some of the less expensive options are either on outdated platforms or wouldn’t have enough core or thread count for the comparisons to make sense. Unfortunately, that leaves us with only two things in inventory. The Core i9 9900K and the Ryzen 9 3900X. If nothing else, this gives you an idea of where the 10980XE falls relative to the top or near the top of the mainstream segment and may be of some use for those deciding between going mainstream or spending the money to step up to the HEDT platform and go all out or at least, go all out on the Intel side.
Additionally, the Cascade Lake-X processors do sit between both company’s mainstream offerings and AMD’s new HEDT parts which are substantially more expensive. Therefore, I think including the 3900X and 9900K for reference makes some sense.
We put the Core i9 10980XE through our testing suite and will show the data below, but a lack of meaningful comparison makes it harder to put the CPU into context. In order to try and do that a bit better, I have also done testing using the Core i9 10980XE with some of its cores disabled to simulate the Core i9 10920X.
Obviously, there is slightly more cache on our sample, and the clocks on the 10920X are slightly better so that must be taken into account. However, it will give you a general idea of what to expect. Furthermore, we did learn some interesting facts about the 10980XE doing so, which I will explain as we go through the data points.
Ultimately means is that we aren’t done on the topic. We will circle back around to the performance when we fill our test bench out with more CPUs. The sad truth is, we have to work with what we have sometimes.