[concurrency-interest] Some interesting (confusing?) benchmark results

Kirk Pepperdine kirk at kodewerk.com
Tue Feb 12 16:52:52 EST 2013


On 2013-02-12, at 10:24 PM, Vitaly Davidovich <vitalyd at gmail.com> wrote:

> For pure CPU bound work, I typically add 1 or maybe 2 more threads than # of hardware threads; this is to account for hardware threads possibly hitting a hard page fault and getting suspended.  I don't see how having any more than that threads benefits perf.
> 
> 

+1,

if you only 60-70% on a core you've got something bad happening. i've always been able to get close to 100% on what should CPU bound jobs when I get rid of bad stuff... that would include a lot of safe-pointing that goes on in the JVM.

Regards,
Kirk

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://cs.oswego.edu/pipermail/concurrency-interest/attachments/20130212/04d55c68/attachment.html>


More information about the Concurrency-interest mailing list