OK, thanks! So I do not have to be worried if I have time to wait :-).
My box is 900^3, so I cannot increase the number of threads when using
GTX1080Ti's.....
bw Dieter
------------------------------------------------------------------------
Dieter Blaas,
Max Perutz Laboratories
Medical University of Vienna,
Inst. Med. Biochem., Vienna Biocenter (VBC),
Dr. Bohr Gasse 9/3,
A-1030 Vienna, Austria,
Tel: 0043 1 4277 61630,
Mobile: 0043 699 1942 1659
Fax: 0043 1 4277 9616,
e-mail: [log in to unmask]
------------------------------------------------------------------------
Am 10.12.2019 um 15:12 schrieb Takanori Nakane:
> Hi,
>
> This depends on many things. How good your CPU is,
> the number of threads and processes assigned to a GPU,
> the number of pooled particles, box size, resolution,
> how easy to orient your particles, etc.
>
> If your box is small enough, you can run two MPI processes
> per GPU to improve GPU utilization.
>
> Best regards,
>
> Takanori Nakane
>
>> Hi,
>>
>> ???? I was just wondering whether the extremely low 'Volatile GPU-Util'
>> values reported by 'nvidia-smi' during refinement are normal. With
>> increasing iteration numbers I see an increase in projected time from
>> about 1 hr to 5 hrs per iteration and out of all four GPUs just one
>> shows 'Volatle GPU-Util' values around 27 % from time to time, the
>> others remain zero. All four GPUs consume about 80 W power with 100%
>> memory usage. Is this the normal behavior? If not, what is wrong?
>>
>> Thanks for explanations,
>>
>> Dieter
>>
>>
>> ------------------------------------------------------------------------
>> Dieter Blaas,
>> Max Perutz Laboratories
>> Medical University of Vienna,
>> Inst. Med. Biochem., Vienna Biocenter (VBC),
>> Dr. Bohr Gasse 9/3,
>> A-1030 Vienna, Austria,
>> Tel: 0043 1 4277 61630,
>> Mobile: 0043 699 1942 1659
>> Fax: 0043 1 4277 9616,
>> e-mail: [log in to unmask]
>> ------------------------------------------------------------------------
>>
>> ########################################################################
>>
>> To unsubscribe from the CCPEM list, click the following link:
>> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1
>>
>
########################################################################
To unsubscribe from the CCPEM list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1
|