The Incredible Performance & Power Efficiency Of AMD Zen 1 vs. Zen 4C

Written by Michael Larabel in Processors on 25 January 2024 at 03:00 PM EST. Page 7 of 7. 33 Comments.
Xmrig benchmark with settings of Variant: Monero, Hash Count: 1M. EPYC 8534PN - Zen 4C was the fastest.
Xmrig benchmark with settings of Variant: Wownero, Hash Count: 1M. EPYC 8534PN - Zen 4C was the fastest.
Xmrig benchmark with settings of Variant: Wownero, Hash Count: 1M. EPYC 8534PN - Zen 4C was the fastest.
Xmrig benchmark with settings of Variant: Wownero, Hash Count: 1M. EPYC 8534PN - Zen 4C was the fastest.
Xmrig benchmark with settings of Variant: CryptoNight-Femto UPX2, Hash Count: 1M. EPYC 8534PN - Zen 4C was the fastest.

By now it should be quite clear about the enormous difference to raw performance and power efficiency going from Zen 1 to Zen 4C. But if you want even more numbers... For this comparison I ran 150+ benchmarks in total for this fresh Naples vs. Siena benchmarking on Ubuntu 23.10 Linux. Go check out all those performance and power efficiency numbers if interested.

Geometric Mean Of All Test Results benchmark with settings of Result Composite, AMD EPYC Zen 4C Power Efficiency. EPYC 8534PN - Zen 4C was the fastest.

When taking the geo mean of all 151 benchmarks conducted across the Naples and Siena processors on Ubuntu 23.10 + Linux 6.6, the difference becomes extremely clear. The EPYC 8324PN NEBS-friendly model at the same 32-cores / 64-threads of the EPYC 7601 was at 2.05x from that original AMD EPYC server processor. Or going to the AMD EPYC 8534P was 2.34x the performance of the EPYC 7601, again at the same core/thread count. Or if going up the stack to the EPYC 8534PN at 64-cores / 128-threads is a 2.87x difference over the original Zen 1 flagship.

CPU Power Consumption Monitor benchmark with settings of Phoronix Test Suite System Monitoring.

When looking at the CPU power consumption over the entire span of benchmarks conducted, the Zen 4C processors really stand out... The EPYC 7601 on average was consuming 143 Watts while the EPYC 8324P was at an 88 Watt average and the EPYC 8342PN at a 71 Watt average. The CPUs were staying below their default TDPs with all of the PowerCap/RAPL monitoring. Even the EPYC 8534PN 64-core processor was on average consuming much less power than the EPYC 7601. It's also worth reiterating the idle difference during the brief periods of downtime between benchmarks... The Siena processors could dip down to just 6~11 Watts compared to Naples that had a minimum power draw of 33 Watts but typically was consuming at least 60 Watts.

The change in AMD EPYC performance and power efficiency from Zen 1 to Zen 4C is outstanding. If you are still on an AMD EPYC 7601 series platform it's really time for an upgrade... Going for AMD EPYC 9004 Genoa(X) / Bergamo allows for even more outstanding results beyond what was shown in this article, but if you are running a 5~7 year server chances are you are really after maximizing your total cost of ownership (TCO) and that is where Siena really spells out a beautiful story. The AMD EPYC 8004 series continue to prove themselves very capable for delivering great performance at unbeatable power efficiency, the lower power consumption leads to reduced platform and cooling/operating costs, and the six memory channels still provide ample DDR5 memory bandwidth while lowering platform costs. The power efficiency of the AMD EPYC 8004 series also opens up EPYC to more deployments outside of the traditional data center setting for more intelligent edge and teclo type scenarios.

So if you were curious like I about how far EPYC performance has come from Zen 1 to Zen 4C, hopefully you found these numbers interesting -- but if you really are still relying on EPYC 7601 series in production, it really is past time to consider upgrading. If interested in any other historical comparisons and fun/interesting benchmark pieces, you can always let me know in the forums or via Twitter/X.

If you enjoyed this article consider joining Phoronix Premium to view this site ad-free, multi-page articles on a single page, and other benefits. PayPal or Stripe tips are also graciously accepted. Thanks for your support.


Related Articles
About The Author
Michael Larabel

Michael Larabel is the principal author of Phoronix.com and founded the site in 2004 with a focus on enriching the Linux hardware experience. Michael has written more than 20,000 articles covering the state of Linux hardware support, Linux performance, graphics drivers, and other topics. Michael is also the lead developer of the Phoronix Test Suite, Phoromatic, and OpenBenchmarking.org automated benchmarking software. He can be followed via Twitter, LinkedIn, or contacted via MichaelLarabel.com.