# hardware bench **URL:** \ **Category:** General\ **Created:** [May 1, 2018, 11:52am UTC](https://forums.chaos.com/t/hardware-bench/88432 "2018-05-01T11:52:25Z")\ **Posts on this page:** 20\ **Page:** 1
**Author:** ![Paul\_Oblomov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/paul_oblomov/32/41296_2.png) [@Paul\_Oblomov](https://forums.chaos.com/u/Paul_Oblomov)\ **Post date:** [May 1, 2018, 11:52am UTC](https://forums.chaos.com/t/hardware-bench/88432/1 "2018-05-01T11:52:25Z")
Based on **car tire burnout** from example scenes. > **[Phoenix for 3ds Max](https://documentation.chaos.com/space/PHX4MAX)** but please, turn off read cache preview during sim (preview tab). used nightly from may 1st. and here’s my results: ```auto CPU INTEL XEON 2699 V4 ES@2.4Ghz [44 threads] RAM Kingston HyperX Savage DDR4 [HX424C14SBK4/64]@14-14-14-35-CR1 Total elapsed 11m 15s Performance 3.49 vox/s ``` P.S. RAM performance tests used: [http://www.softpedia.com/get/System/Benchmarks/MaxxMEM2.shtml https://www.aida64.com/downloads](http://www.softpedia.com/get/System/Benchmarks/MaxxMEM2.shtml) ![MaxxMEM2 results.jpg](https://europe1.discourse-cdn.com/chaosprd/original/3X/8/2/82fbce64f9fc75327806a28227c931ed4f5441d0.jpeg) ![AIDA64 extreme memory test.png](https://europe1.discourse-cdn.com/chaosprd/original/3X/e/a/eadd6386d532ae7cc82e0d6c7d7970c7f0855396.png) ---
**Author:** ![Paul\_Oblomov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/paul_oblomov/32/41296_2.png) [@Paul\_Oblomov](https://forums.chaos.com/u/Paul_Oblomov)\ **Post date:** [May 1, 2018, 1:35pm UTC](https://forums.chaos.com/t/hardware-bench/88432/2 "2018-05-01T13:35:25Z")
Heh. Not bad. I wonder, where all the AMD ryzen or epycs ? ---
**Author:** ![jacksc02](https://avatars.discourse-cdn.com/v4/letter/j/779978/32.png) [@jacksc02](https://forums.chaos.com/u/jacksc02)\ **Post date:** [August 14, 2018, 8:53am UTC](https://forums.chaos.com/t/hardware-bench/88432/3 "2018-08-14T08:53:01Z")
I will try it out tomorrow on 2990wx ---
**Author:** ![Paul\_Oblomov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/paul_oblomov/32/41296_2.png) [@Paul\_Oblomov](https://forums.chaos.com/u/Paul_Oblomov)\ **Post date:** [August 14, 2018, 9:09am UTC](https://forums.chaos.com/t/hardware-bench/88432/4 "2018-08-14T09:09:47Z")
Btw, would be great to have phx logs also. Interesting what would be the slowest part. ---
**Author:** ![Svetlin.Nikolov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/svetlin.nikolov/32/64983_2.png) [@Svetlin.Nikolov](https://forums.chaos.com/u/Svetlin.Nikolov)\ **Post date:** [August 14, 2018, 9:35pm UTC](https://forums.chaos.com/t/hardware-bench/88432/5 "2018-08-14T21:35:34Z")
> [@jacksc02](#): > > I will try it out tomorrow on 2990wx Threadrippers are unfortunately NUMA inside, so they won’t perform well. You can check here for more info on the recommended hardware: [Chaos Docs](https://docs.chaosgroup.com/display/PHX3MAX/Setup+and+Installation#SetupandInstallation-HardwareAdviceforFasterSimulating) Cheers! ---
**Author:** ![Raven](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/raven/32/12719_2.png) [@Raven](https://forums.chaos.com/u/Raven)\ **Post date:** [August 15, 2018, 1:54pm UTC](https://forums.chaos.com/t/hardware-bench/88432/6 "2018-08-15T13:54:20Z")
@Svetlin So I know there are many factors involved as per the recommendation document. What machines do you guys prefer to use? I’m looking at a new machine soon, and I can get the i9-7900 or a Threadripper 1950x for the same price. Now I don’t do tons of fluid work, but enough that I want to consider it in my setup. If I understand the NUMA issue correctly, the threadripper would effectively be an 8 core machine vs the i9s 10 cores. Thanks ---
**Author:** ![Paul\_Oblomov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/paul_oblomov/32/41296_2.png) [@Paul\_Oblomov](https://forums.chaos.com/u/Paul_Oblomov)\ **Post date:** [August 15, 2018, 2:15pm UTC](https://forums.chaos.com/t/hardware-bench/88432/7 "2018-08-15T14:15:00Z")
Let us see TR results first 🙂 ---
**Author:** ![Svetlin.Nikolov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/svetlin.nikolov/32/64983_2.png) [@Svetlin.Nikolov](https://forums.chaos.com/u/Svetlin.Nikolov)\ **Post date:** [August 15, 2018, 6:08pm UTC](https://forums.chaos.com/t/hardware-bench/88432/8 "2018-08-15T18:08:02Z")
@georgi.zhekov should be able to say more specifically what model of TR we tested Phoenix on - it was a 16 core, 32 thread CPU and produced roughly the same results as an i7-5820K, except for some (not all) FLIP scenes where the TR is slower. We need to check the bus speed on both configurations though, because it is just as important as the CPU. Using a NUMA CPU won’t automatically exclude half the cores - you would need to do it manually from the Simulation rollout. However, with Threadrippers and some Xeons that are NUMA inside but don’t register as such in Windows, this can’t be done because as much as Phoenix could gather from the OS is that it’s a plain old CPU. You could try to work around this by setting the thread number to half the threads and hope that Windows orders the first NUMA node cores first and the rest second (for a 2-NUMA config). Otherwise, using all cores on a NUMA won’t automatically halve the speed. The sim could work a bit slower or even more than twice slower, depending on the CPU. ---
**Author:** ![georgi.zhekov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/georgi.zhekov/32/5768_2.png) [@georgi.zhekov](https://forums.chaos.com/u/georgi.zhekov)\ **Post date:** [August 16, 2018, 8:47am UTC](https://forums.chaos.com/t/hardware-bench/88432/9 "2018-08-16T08:47:59Z")
The TR we tested was the 1950x vs the i7-5820K and as Svetlin said it was pretty much the same result, although it has quite a lot more cores. The bus speed is around the same for both CPUs. ---
**Author:** ![Raven](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/raven/32/12719_2.png) [@Raven](https://forums.chaos.com/u/Raven)\ **Post date:** [August 16, 2018, 9:40am UTC](https://forums.chaos.com/t/hardware-bench/88432/10 "2018-08-16T09:40:38Z")
Thanks for the info guys, So I guess for now the i9 will perform better in the sim department. Is the NUMA issue something that is solvable in the near future, or would it require a full re-look at the core, as I guess that processors like the TR2 will seam like they under perform. Do you know if NUMA affects other simulators ie Houdini or realflow? ---
**Author:** ![nikola.bozhinov](https://avatars.discourse-cdn.com/v4/letter/n/c2a13f/32.png) [@nikola.bozhinov](https://forums.chaos.com/u/nikola.bozhinov)\ **Post date:** [August 20, 2018, 10:36am UTC](https://forums.chaos.com/t/hardware-bench/88432/11 "2018-08-20T10:36:21Z")
Hi, The problems with the NUMA configurations (and the distributed simulations) is inherent to some of the basic algorithms of the fluid simulations, so solving them is not just implementation problem. This doesn’t mean that no optimizations are possible, but it is unclear how much things can be improved with the current algorithms. In general fluid dynamics simulations are more memory intensive than rendering and that is why multi-core and especially NUMA systems seem to under perform - the bus speed is often the limiting factor, as [noted in the docs](https://docs.chaosgroup.com/display/PHX3MAX/Setup+and+Installation#SetupandInstallation-HardwareAdviceforFasterSimulating). And in the NUMA processors the different cores need different time to access the memory, so the problem becomes even greater there. Have in mind that some processors with high core count have NUMA architecture even if that is not said explicitly. ---
**Author:** ![Paul\_Oblomov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/paul_oblomov/32/41296_2.png) [@Paul\_Oblomov](https://forums.chaos.com/u/Paul_Oblomov)\ **Post date:** [August 22, 2018, 6:59am UTC](https://forums.chaos.com/t/hardware-bench/88432/12 "2018-08-22T06:59:56Z")
I wonder, why still there’s no TR benches here ? What’s happened ?! ---
**Author:** ![Paul\_Oblomov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/paul_oblomov/32/41296_2.png) [@Paul\_Oblomov](https://forums.chaos.com/u/Paul_Oblomov)\ **Post date:** [October 4, 2018, 1:03am UTC](https://forums.chaos.com/t/hardware-bench/88432/13 "2018-10-04T01:03:22Z")
I saw vlado got his hands on tr 2990. Plz can you test itat full powe and at maybe one memblock out of four (32/4) cores? ---
**Author:** ![jacksc02](https://avatars.discourse-cdn.com/v4/letter/j/779978/32.png) [@jacksc02](https://forums.chaos.com/u/jacksc02)\ **Post date:** [October 4, 2018, 10:08am UTC](https://forums.chaos.com/t/hardware-bench/88432/14 "2018-10-04T10:08:09Z")
2990wx Total elapsed time - 14m 31s ---
**Author:** ![Paul\_Oblomov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/paul_oblomov/32/41296_2.png) [@Paul\_Oblomov](https://forums.chaos.com/u/Paul_Oblomov)\ **Post date:** [October 4, 2018, 11:05am UTC](https://forums.chaos.com/t/hardware-bench/88432/15 "2018-10-04T11:05:50Z")
Weee 🙂 My old sh\*t is still capable 😄 Thx! ---
**Author:** ![Paul\_Oblomov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/paul_oblomov/32/41296_2.png) [@Paul\_Oblomov](https://forums.chaos.com/u/Paul_Oblomov)\ **Post date:** [October 10, 2018, 10:36pm UTC](https://forums.chaos.com/t/hardware-bench/88432/16 "2018-10-10T22:36:17Z")
Installed new win10 1809 - and my results are 13minutes11seconds (3.27mvox). I turned off meltdown updates and still the same. Vraybench renders with the same speed. Interesting… ---
**Author:** ![3dwolf](https://avatars.discourse-cdn.com/v4/letter/3/ecae2f/32.png) [@3dwolf](https://forums.chaos.com/u/3dwolf)\ **Post date:** [October 16, 2018, 10:51am UTC](https://forums.chaos.com/t/hardware-bench/88432/17 "2018-10-16T10:51:44Z")
```auto Dual Intel Xeon E5-2650 v2 … 2x 8 cores Total Elapsed: Both Numa. 17m6s Single Numa: 23m15s ``` EDIT: 14m6s on 3.12 → indeed ~20% faster 🕶 ![NUMA_both.JPG](https://europe1.discourse-cdn.com/chaosprd/original/3X/8/0/80f808c5a3e19c0e765e9bb0ef715d11c7d684bd.jpeg) ![NUMA_only1.JPG](https://europe1.discourse-cdn.com/chaosprd/original/3X/f/8/f858226178ae5570b87550715c2679ba5a32cee1.jpeg) ![MaxxMEM2_c3249.jpg](https://europe1.discourse-cdn.com/chaosprd/original/3X/3/e/3e0f15877d9c750e5c949b7c81f5bb0739c4ed74.jpeg) ![AIDA64.JPG](https://europe1.discourse-cdn.com/chaosprd/original/3X/0/f/0ffc9648bee5f215d5c8c83352f45405358a89ab.jpeg) ---
**Author:** ![Svetlin.Nikolov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/svetlin.nikolov/32/64983_2.png) [@Svetlin.Nikolov](https://forums.chaos.com/u/Svetlin.Nikolov)\ **Post date:** [October 16, 2018, 11:06am UTC](https://forums.chaos.com/t/hardware-bench/88432/18 "2018-10-16T11:06:52Z")
Hey, Make sure to grab a latest nightly - we have some large test scenes here that are running 20-30% faster simulations now compared to 3.04 🙂 Cheers! ---
**Author:** ![georgi.zhekov](https://dub1.discourse-cdn.com/chaosprd/user_avatar/forums.chaos.com/georgi.zhekov/32/5768_2.png) [@georgi.zhekov](https://forums.chaos.com/u/georgi.zhekov)\ **Post date:** [October 20, 2018, 3:15pm UTC](https://forums.chaos.com/t/hardware-bench/88432/19 "2018-10-20T15:15:57Z")
Running the scene on Ryzen 2700x - 13m 51s Using a nightly build from 20.10.2018 🙂 ---
**Author:** ![waaazoo](https://avatars.discourse-cdn.com/v4/letter/w/a9adbd/32.png) [@waaazoo](https://forums.chaos.com/u/waaazoo)\ **Post date:** [July 9, 2019, 8:56am UTC](https://forums.chaos.com/t/hardware-bench/88432/20 "2019-07-09T08:56:04Z")
Up for this already “old” topic. I’m simming on a Threadripper 2950X (DDR4 3200 and M2 drive). Previsouly I was simming on a Threadripper 1950X. I get small speed up. I’m still thinking about a new system, only for Sims, but I’m still wondering what would be a good/better system. Will test the AIDA benchmark just to compare with what we have above 🙂 Still no benchmark tool around for PhoenixFD ? [Next page](https://forums.chaos.com/t/hardware-bench/88432.md?page=2)