Re: [RFC v5 00/11] Speculative page faults

[Balbir Singh]

On Mon, 2017-07-03 at 19:32 +0200, Laurent Dufour wrote:
> The test is counting the number of records per second it can manage, the
> higher is the best. I run it like this 'ebizzy -mTRp'. To get consistent
> result I repeat the test 100 times and measure the average result, mean
> deviation and max. I run the test on top of 4.12 on 2 nodes, one with 80
> CPUs, and the other one with 1024 CPUs:
> 
> * 80 CPUs Power 8 node:
> Records/s	4.12		4.12-SPF
> Average		38941,62	64235,82
> Mean deviation	620,93		1718,95
> Max		41988		69623
> 
> * 1024 CPUs Power 8 node:
> Records/s	4.12		4.12-SPF
> Average		39516,64	80689,27
> Mean deviation	1387,66		1319,98
> Max		43281		90441
>

This seems like a very interesting result

Balbir Singh.