On Mon, Jun 16, 2025 at 09:47:52AM -0700, Eric Biggers wrote:
FWIW, on an AMD EPYC 9B45 (Zen 5 / Turin) server processor, I get 35.2 GB/s. This processor appeared to run at about 4.15 GHz, so that's about 8.5 bytes per cycle. That's 51% more bytes per cycle than Intel. This shows that there is still room for improvement in VAES, even when it's already much better than QAT.
Also, to be clear, the 35.2 GB/s (and the corresponding bytes/cycle number of 8.5) is single-thread throughput. This should have been clear since I compared it to the single-thread throughput on Emerald Rapids. But I just wanted to make sure to state it explicitly, as an earlier part of my email discussed whole-processor throughput which it could be confused with.
- Eric