CERN analyzed performance+power usage for some HPC CPU candidates, with expected results: http://lvalsan.web.cern.ch/lvalsan/processor_benchmarking/presentation/ …pic.twitter.com/336T8pgZR3
-
@marcan42 the new Atoms are a lot better -- still terrible performance-wise, but at least very low power. Broadwells might be better though
@FioraAeterna My old opinion of Atom was "good if you're TDP-limited or need cheap" - Sandy Bridge used to smoke them at perf/W in its day.
-
-
@marcan42@FioraAeterna I expect that Silvermont will lose to Haswell in perf/W for SIMD-friendly code, though. (bcrypt is scalar.) -
@solardiz@FioraAeterna IIRC bcrypt is basically evil random-mem-heavy code (i.e. not optimizable/pipelinable) so I can't say I'm surprised - Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.