AMD 3rd Gen EPYC Milan Review: A Peak vs Per Core Performance Balance
by Dr. Ian Cutress & Andrei Frumusanu on March 15, 2021 11:00 AM ESTDisclaimer June 25th: The benchmark figures in this review have been superseded by our second follow-up Milan review article, where we observe improved performance figures on a production platform compared to AMD’s reference system in this piece.
Compiling LLVM, NAMD Performance
As we’re trying to rebuild our server test suite piece by piece – and there’s still a lot of work go ahead to get a good representative “real world” set of workloads, one more highly desired benchmark amongst readers was a more realistic compilation suite. Chrome and LLVM codebases being the most requested, I landed on LLVM as it’s fairly easy to set up and straightforward.
git clone https://github.com/llvm/llvm-project.gitcd llvm-projectgit checkout release/11.xmkdir ./buildcd ..mkdir llvm-project-tmpfssudo mount -t tmpfs -o size=10G,mode=1777 tmpfs ./llvm-project-tmpfscp -r llvm-project/* llvm-project-tmpfscd ./llvm-project-tmpfs/buildcmake -G Ninja \ -DLLVM_ENABLE_PROJECTS="clang;libcxx;libcxxabi;lldb;compiler-rt;lld" \ -DCMAKE_BUILD_TYPE=Release ../llvmtime cmake --build .We’re using the LLVM 11.0.0 release as the build target version, and we’re compiling Clang, libc++abi, LLDB, Compiler-RT and LLD using GCC 10.2 (self-compiled). To avoid any concerns about I/O we’re building things on a ramdisk. We’re measuring the actual build time and don’t include the configuration phase as usually in the real world that doesn’t happen repeatedly.
For the new Milan chips, the results are a bit mixed. The higher-power 7763 takes a lead with a +10.5% improvement over the 7742, however the 7713 doesn’t manage to keep up with that predecessor.
The 1S vs 2S scores are interesting as the 2S figures showcase the new Milan chips in a better light due to the higher single-threaded performance of the Zen3 cores. The compilation here also has linking phases which are single-thread performance bottle-necked. This results in scenarios such as the 7713 losing to the 7662 in 1S comparisons, however winning out against the same chip in the 2S comparison, as it’s able to make that advantage count for more.
It’s also great to see the 75F3 keep up with the 64-core counterparts at around 72% of the top-SKU performance.
Finally, in NAMD, this is more of a core-local compute workload. We see the 7763 outperform the 7742 by +11.8%, however the Milan chip is still outperformed by the higher core compute capacity of the 80-core Altra chip.
Generally, I have my reservations about NAMD as a benchmark due to its multicore vs MPI variants and scaling anomalies, on top of the whole topic of the benchmark having a completely different algorithm for AVX512 processors.
Facebook
Twitter
RSS
120 Comments
View All Comments
mode_13h - Monday, March 15, 2021 - link
Please don't paint Nvidia as a victim. They are not. All of these guys will have to support each other, for the foreseeable future, and for purely pragmatic reasons.Oxford Guy - Monday, March 15, 2021 - link
They are not 'guys'. They're corporations. Corporations were invented to, to quote Ambrose Bierce, grant 'individual profit without individual responsibility'.mode_13h - Wednesday, March 17, 2021 - link
No disagreement, but I'm slightly disheartened you decided to take issue with my use of the term "guys". I'll try harder, next time--just for you.Oxford Guy - Tuesday, April 6, 2021 - link
People humanize corporations all the time. It doesn't lead to good outcomes for societies.Of course, it's questionable whether corporations lead to good outcomes, considering that they're founded on scamming people (profit being 'sell less for more', needing tricks to get people to agree to that).
chavv - Monday, March 15, 2021 - link
Is it possible to add another "benchmark" - ESX server workload?Like, running 8-16-32-64 VMs all with some workload...
Andrei Frumusanu - Monday, March 15, 2021 - link
As we're rebuilding our server test suite, I'll be looking into more diverse benchmarks to include. It's a long process that needs a lot of thought and possibly resources so it's not always evident to achieve.eva02langley - Monday, March 15, 2021 - link
Just buy EPYC and start your hybridation and your reliance on a SINGLE supplier...eva02langley - Monday, March 15, 2021 - link
edit: Just buy EPYC and start your hybridation and STOP your reliance on a SINGLE supplier...mode_13h - Monday, March 15, 2021 - link
You guys should really include some workloads involving multiple <= 16-core/32-thread VMs, that could highlight the performance advantages of NPS4 mode. Even if all you did was partition up the system into smaller VMs running multithreaded SPEC 2017 tests, at least that would be *something*.That said, please don't get rid of all system-wide multithreaded tests, because we definitely still want to see how well these systems scale (both single- and multi- CPU).
ishould - Monday, March 15, 2021 - link
Yes this seems more useful for my needs as well. We use a grid system for job submission and not all cores will be hammered at the same time