AMD’s New EPYC 7F52 Reviewed: The F is for ᴴᴵᴳᴴ Frequency
by Dr. Ian Cutress on April 14, 2020 9:45 AM EST- Posted in
- CPUs
- AMD
- Enterprise
- Enterprise CPUs
- EPYC
- SP3r2
- CPU Frequency
- Rome
- 7Fx2
SPEC2006 and SPEC2017 (Single Thread)
Due to some limitations with our systems, we were only able to run SPEC in single thread mode in time for the review. Given that these 7F processors are meant to be the highest frequency EPYC hardware available, in single thread and multi-thread, this is still a very relevant test for the use case. Unfortunately we introduced this test late last year, after testing the bulk of our Intel CPUs. We’re currently re-running on a few and will update this post over the next few days.
*If you are seeing this as the review goes live, we are still waiting for the 6226R results to finish.
SPEC2006 1T Estimated Results | ||||||||
AnandTech | AMD 7F52 |
AMD 7601 |
AMD 3990X |
AMD 3950X |
Intel 6226R |
Intel 9900KS |
Intel 10980XE |
|
uArch | Rome | Naples | Rome | Rome | CLX-R | Coffee | CLX | |
Turbo | 3900 | 3200 | 4300 | 4700 | 3900 | 5000 | 4800 | |
400.perlbench | 45.9 | 29.8 | 50.8 | 54.6 | 40.2 | 60.1 | 55.2 | |
401.bzip2 | 30.9 | 23.3 | 34.5 | 36.6 | 25.4 | 37.5 | 33.5 | |
403.gcc | 37.7 | 28.0 | 53.4 | 57.7 | 30.0 | 56.1 | 46.6 | |
429.mcf | 35.6 | 22.6 | 48.6 | 52.9 | 28.5 | 64.7 | 45.3 | |
445.gobmk | 36.7 | 23.4 | 41.8 | 44.9 | 32.0 | 43.3 | 39.6 | |
456.hmmr | 36.8 | 26.8 | 41.0 | 43.3 | 39.2 | 51.7 | 48.2 | |
458.sjeng | 32.5 | 21.9 | 38.1 | 41.1 | 34.7 | 47.0 | 43.6 | |
462.libquantum | 78.7 | 50.3 | 100.4 | 102.8 | 38.5 | 113.2 | 106.8 | |
464.h264ref | 67.7 | 49.6 | 75.9 | 80.4 | 64.7 | 83.9 | 79.1 | |
471.omnetpp | 21.1 | 14.0 | 27.5 | 31.9 | 25.5 | 31.3 | 30.0 | |
473.astar | 26.9 | 17.8 | 30.9 | 32.8 | 22.9 | 30.2 | 29.5 | |
483.xalancbmk | 46.0 | 29.2 | 53.8 | 58.0 | 37.5 | 60.4 | 54.6 | |
433.milc | 35.0 | 22.6 | 46.9 | 49.3 | 15.7 | 31.9 | 27.9 | |
444.namd | 39.0 | 29.6 | 43.3 | 45.9 | 38.3 | 52.5 | 43.9 | |
450.soplex | 58.9 | 39.7 | 73.7 | 74.8 | 21.5 | 73.0 | 67.1 | |
453.povray | 59.7 | 37.0 | 66.3 | 70.9 | 58.5 | 76.2 | 70.5 | |
470.lbm | 101.4 | 72.4 | 121.8 | 126.2 | 20.2 | 77.7 | 102.9 | |
482.sphinx3 | 94.7 | 56.2 | 107.4 | 113.0 | 45.3 | 105.0 | 72.6 | |
Geomean | 44.8 | 30.2 | 53.6 | 57.1 | 32.3 | 56.6 | 51.1 |
The performance jump from the Naples 7601 to the Rome 7F52 is bordering on about 50%. It is worth pointing out that AMD’s consumer Ryzen 9 3950X wins out here due to IPC and single core frequency, closely followed by Intel’s i9-9900KS, the AMD Threadripper 3000s, and the Intel i9-10980XE. This comes down to consumer platforms affording much larger turbos and not being stricter on RAS requirements and such.
SPEC2017 1T Estimated Results | ||||||||
AnandTech | AMD 7F52 |
AMD 7601 |
AMD 3990X |
AMD 3950X |
Intel 6226R |
Intel 9900KS |
Intel 10980XE |
|
uArch | Rome | Naples | Rome | Rome | CLX-R | Coffee | CLX | |
Turbo | 3900 | 3200 | 4300 | 4700 | 3900 | 5000 | 4800 | |
500.perlbench_r | 4.3 | 2.7 | 5.0 | 5.3 | 5.1 | 6.9 | 6.3 | |
502.gcc_r | 6.1 | 4.4 | 8.0 | 8.6 | 3.8 | 9.3 | 7.4 | |
505.mcf_r | 5.0 | 3.5 | 6.1 | 6.6 | 3.2 | 6.5 | 5.4 | |
520.omnetpp_r | 2.4 | 2.0 | 3.4 | 3.7 | 3.1 | 4.1 | 3.8 | |
523.xalancbmk_r | 4.7 | 2.5 | 5.0 | 5.3 | 4.0 | 4.4 | 5.3 | |
525.x264_r | 7.8 | 5.7 | 9.0 | 9.5 | 6.8 | 9.7 | 9.0 | |
531.deepsjeng_r | 3.7 | 3.0 | 4.4 | 4.7 | 4.0 | 5.5 | 5.0 | |
541.leela_r | 4.1 | 2.9 | 4.6 | 4.9 | 3.7 | 5.0 | 4.6 | |
548.exchange2_r | 7.3 | 4.5 | 8.2 | 8.7 | 6.2 | 8.3 | 7.5 | |
557.xz_r | 3.0 | 2.1 | 3.8 | 4.1 | 2.9 | 4.1 | 3.8 | |
503.bwaves_r | 39.7 | 27.4 | 46.5 | 48.5 | 7.4 | 38.2 | 30.6 | |
507.cactuBSSN_r | 5.6 | 4.2 | 6.4 | 6.7 | 4.3 | 8.3 | 6.1 | |
508.namd_r | 6.0 | 4.6 | 6.7 | 7.0 | 4.1 | 7.4 | 6.3 | |
510.parest_r | 7.5 | 5.5 | 8.4 | 9.0 | 4.4 | 9.7 | 7.4 | |
511.povray_r | 6.7 | 4.2 | 7.5 | 7.9 | 6.6 | 8.7 | 8.0 | |
519.lbm_r | 6.9 | 5.0 | 8.0 | 8.4 | 1.0 | 7.7 | 6.3 | |
521.wrf_r * | - | - | - | - | - | - | - | |
526.blender_r | 6.6 | 4.7 | 7.5 | 8.0 | 5.2 | 7.9 | 7.2 | |
527.cam4_r | 6.8 | 4.8 | 7.7 | 8.2 | 4.8 | 8.3 | 6.4 | |
538.imagick_r | 7.9 | 5.8 | 8.8 | 9.4 | 6.4 | 8.5 | 7.8 | |
544.nab_r | 4.0 | 3.0 | 4.4 | 4.7 | 3.0 | 5.2 | 4.7 | |
549.fotonik_r | 14.2 | 8.1 | 17.2 | 16.4 | 3.5 | 14.8 | 11.4 | |
554.roms_r | 9.0 | 5.3 | 10.9 | 11.4 | 3.8 | 10.0 | 7.3 | |
Geomean | 6.3 | 4.3 | 7.3 | 7.7 | 4.1 | 7.8 | 6.8 | |
*512.wrf_r unfortunately doesn't run properly in our SPEC harness at this time |
We see a similar result in the newer version of SPEC, again with ~50% jump from the Naples 7601 to the Rome 7F52. The 9900KS has the overall better Geomean here, followed closely behind by the 3950X, then the Threadrippers.
97 Comments
View All Comments
beginning - Tuesday, April 14, 2020 - link
Thank you for this review! Quite useful. I would like to see performance comparisons on multi-threaded workloads too.anonomouse - Tuesday, April 14, 2020 - link
Looking at memory latency chart, it doesn't look like the second half of the LLC is actually any farther - that's just the limit of the range that the L2 TLB (2K entries @4K pages) can cover. The additional latency beyond 8MB is probably because requests are now causing table walks, and that is also clear from Andrei's corresponding chart for the client version in 3700x/3900x. Were these results cross-checked with each other before publication?One way to verify this would be to enable huge pages and see if the memory latency profile you're claiming looks different.
boozed - Tuesday, April 14, 2020 - link
"AMD is hoping to stag that first second rung of the ladder"What does this expression mean?
UnknownKnolwdge - Tuesday, April 14, 2020 - link
Look up ladder logic on wikipedia, search for rung.It's associated with relay logic and old style PLC programming.
boozed - Tuesday, April 14, 2020 - link
Thanks but I'm not convinced that's itSmell This - Wednesday, April 15, 2020 - link
"AMD is hoping to stag that first second rung of the ladder."________________________________________________
Two things:
1) I suspect the meaning is 'the next step up;' and
2) 'snag' instead of 'stag'
:-)
madwolfchin - Tuesday, April 14, 2020 - link
This processor should be targeted for Single threaded workload, why are all the benchmark multi threaded. Would also like to see the IPC performance since there is a large increase of cachenicamarvin - Wednesday, April 15, 2020 - link
Mr. Cutress on this article you have the 3950X and 3990X as "Rome" uArch, but in Fact they are Matisse for 3950X and Castle Peak for the 3990XMachinus - Wednesday, April 15, 2020 - link
I wish there was more than one CPU maker in the market right now. These chips looks great, but there is no competition.Threska - Wednesday, April 15, 2020 - link
Not specifically in this niche, but ARM (and equivalent) haven't gone anywhere.