If I wanted to do a pure serial vs. parallel comparison in Chapel, I'd
change my 'forall' loops to 'for' loops. For a long-running loop, it's
quite likely that you'd get similar results with either approach, but
there's a bunch of machinery underneath that serial + forall +
dataParTasksPerLocale combination that I'd prefer to completely factor out
of such comparisons to avoid confusion or surprises.
-Brad
On Thu, 23 Jul 2020, Damian McGuckin wrote:
Hi all.
I am doing some benchmarks of some code between its serial performance and
parallel performance.
Is there much overhead in running a forall subjected to a serial clause.
serial
{
....
forall i .... do
{
....
}
}
I also use
--dataParTasksPerLocale=1
for good measure.
Or, for accurate serial performance testing, should I convert my 'forall'
statements to be 'for'.
Stay safe - Damian
Pacific Engineering Systems International, 277-279 Broadway, Glebe NSW 2037
Ph:+61-2-8571-0847 .. Fx:+61-2-9692-9623 | unsolicited email not wanted here
Views & opinions here are mine and not those of any past or present employer
_______________________________________________
Chapel-developers mailing list
[email protected]
https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.sourceforge.net_lists_listinfo_chapel-2Ddevelopers&d=DwICAg&c=C5b8zRQO1miGmBeVZ2LFWg&r=QUQW-BniEL_d2a7btR4rP5TPiNmpm1pG-Qa_xXzGVKc&m=itaFPwvAkemHGNei6ucyyGSrbBuhSM2gTaT1OLK9rBo&s=HkFUmW4ocA-MsPG_MjyeERwjYFEfeK0-vuh7z16Gqf4&e=
_______________________________________________
Chapel-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/chapel-developers
