On 28.01.2020 10:45, Sven Schreiber wrote:
Am 28.01.2020 um 09:27 schrieb Marcin Błażejowski:
> On 28.01.2020 09:18, Sven Schreiber wrote:
>> But openblas alone doesn't explain the equality of the 4-thread and
>> the single-thread outcome, at least comparing to the Windows results.
>> Or did you compile openblas yourself? (Perhaps with a different dgemm
>> threshold?)
>>
> No, I did not. This is a standard debian package
> (libopenblas0-openmp:amd64), but I'm going to compile it so:
Hm, what's your Debian and/or openblas library version then?
This is 'libopenblas-base:amd64 0.3.7+ds-7' which is transitional to (in
my case) 'libopenblas0:amd64 0.3.7+ds-7' which is meta package and true
one is 'libopenblas0-openmp:amd64 0.3.7+ds-7'.
GEMM_MULTITHREAD_THRESHOLD, see Allin's message on Jan 24th.
Some
people replace the default of 4 with 50, apparently.
Ok, I'll look at this.
Cheers,
Marcin
--
Marcin Błażejowski