On Sat, 19 Aug 2000, Ingo Molnar wrote:
> On Fri, 18 Aug 2000, Matthew Wilcox wrote:
> > (From memory, we were acheiving 80%+ scalability on 2 and 4 CPUs and
> > then dropped to 65% on the 8 CPU test -- equivalent to using 5 CPUs at
> > 100%.)
>
> i can show you pure user-space workloads that will not scale past 2 CPUs
> on any x86 SMP system.
Can you point me/us to any references on the SMP scalability of various
workloads and/or how to diagnose them?
I suppose if you're runing 1/2 as slow as you think you should be on a 4
CPU box and you disable two of the CPUs and it doesn't get any slower then
you know you've got a scalability problem somewhere. I don't know how I
would got about determining that the userspace workload inherantly didn't
scale though.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Wed Aug 23 2000 - 21:00:05 EST