3 ms·
Seems the 1.5x vs EPYC claim is from SPECrate_2017_int_base benchmark.
by fulafel 5y ago
Seems the 1.5x vs EPYC claim is from SPECrate_2017_int_base benchmark.
- hrgiger 5y agoafaik that bench still widely used but the epyc rome they mentioned is zen 2. As i understand it is still slightly faster than milan series but the power consumption is half of it. Here is benchmark result of 7713p https://www.spec.org/cpu2017/results/res2021q4/cpu2017-20210927-29554.html https://www.spec.org/cpu2017/results/res2021q4/cpu2017-20210...
- ip26 5y agoIt seems like 2P Milan would be the suitable comparison, based on core count, wattage, CXL link, and so on.
- hrgiger 5y agoyou are right here 7763 result searchable when selecting processor 870ish in general, btw CXL not there yet https://www.spec.org/cgi-bin/osgresults?conf=cpu2017 https://www.spec.org/cgi-bin/osgresults?conf=cpu2017
- TooKool4This 5y agoAND it seems to be based on a simulation which can be fudged in all sorts of ways. Nvidia’s track record with cpus is spotty at best. Not sure how much weight to put on these claims, at least until there is a shipping product. > Nvidia based this claim on a pre-silicon simulation that predicts the Grace CPU at a score of 740+ (370 per chip).