Dear all,
please forgive me if this question has already come up before:
Running "Refine3D" on two 1070 GPUs and 3 CPUs is quite quick but
maximization takes most of the time as it is not (yet?) implemented for
GPU. So, what would be the best parameters to set in 'Number of MPI
Processes" and "Number of Threads" to have two slave threads each
running one of the GPUs during alignement but more than 3 CPUs (I have
8) during maximization? When setting more than 3 CPUs the RAM of the
GPUs becomes automatically split amongst the threads and quickly runs out.
Thanks, Dieter
------------------------------------------------------------------------
Dieter Blaas,
Max F. 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,
Fax: 0043 1 4277 9616,
e-mail: [log in to unmask]
------------------------------------------------------------------------
|