3 ms·
Have you seen https://www.spec.org/cpu2017/results/res2018q3/cpu2017-20180805-08109.html https://www.spec.org/cpu2017/results/res2018q3/cpu2017-20180... from ba
by tpearson-raptor 8y ago
Have you seen https://www.spec.org/cpu2017/results/res2018q3/cpu2017-20180805-08109.html https://www.spec.org/cpu2017/results/res2018q3/cpu2017-20180... from back in July? There are more submissions as well if you want to browse around a bit.
- sliken 8y agoI looked around on the floating point, and found none. There are some integer throughput results. Is the power9 only competitive for integer related workloads? E950 finds results on the Integer throughput list, but not the FP throughput list.
- sliken 8y agoAh, looks like someone formatted the IBM entries incorrectly, at least unlike all the AMD and Intel entries I could find. Usually the description has a CPU entry something like "Intel Xeon Gold 6148". The IBM entries have things like "3.4 - 3.8 GHz, 40 core, SLES". So if you search any of the lists for "power9" you don't find anything.
- sliken 8y agoPhoronix just did a comparison. Power9 22 core, did well on 7zip, octave, and rust, but lost most of the benchmarks to the Xeon and Epyc dual sockets. Seems like Power9 does really well on interpreters.