Re: [ccp4bb] Relion's own implementation of motioncorr2 on gpu cluster

[mesters]

Hi,

the relion BB can be reached at cc...@jiscmail.ac.uk

To answer your question:

There are several implementations / software-packages for setting up an 
MPI on linux and if done correctly, Relion can be run on any of them 
(assuming MPI and Relion are implemented on the cluster). Login to the 
cluster and type "mpirun" in a terminal to test if the software is 
installed. Have a look at the Relion Wiki for the way to run jobs on a 
cluster:


https://www3.mrc-lmb.cam.ac.uk/relion/index.php/Benchmarks_%26_computer_hardware#Assemble_your_own_GPU_machines

Running Relion jobs on a cluster however is not going to get the job 
done faster as the main bottlenecks are
 1) the dataset-containing HDD that will need to feed all the nodes 
with massive amounts of data (TBs rather than GBs),

 2) the speed (10G would be the minimum) & load of the network
and
 3) cluster limitations like time, nodes allocated to one job and 
overall load of the cluster

may all have a dramatic impact on the time required to get the job done.

A dedicated stand alone maschine with a modern 8-12 core processor, a 
minimum of 128MB Ram and 1 or 2 NVIDIA RTX (8GB of Ram each) will do the 
job faster. See the benchmarks on the Relion Wiki page mentioned above.


Cheers,

Jeroen




Am 11.09.20 um 01:18 schrieb Lora Sinha:
In the Relion 3.1 tutorial, the job is submitted to a local machine, 
but I wonder if I can submit the job to a GPU cluster? What MPI 
configuration should I use?





To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1



--
*Dr.math. et dis. nat.Jeroen R. Mesters*
Deputy, Lecturer, Program Coordinator /Infection Biology
/ 
Visiting 
Professorship (South Bohemian University) in Biophysics

Logo Uni Lübeck
*University of Lübeck*
Center for Structural and Cell Biology in Medicine
*Institute of Biochemistry*

Tel +49 451 3101 3105 (secretariate 3101)
Fax +49 451 3101 3104
jeroen.mest...@uni-luebeck.de 
www.biochem.uni-luebeck.de 

*Ratzeburger Allee 160
23538 Lübeck, Schleswig-Holstein
Germany*




To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1

This message was issued to members of www.jiscmail.ac.uk/CCP4BB, a mailing list 
hosted by www.jiscmail.ac.uk, terms & conditions are available at 
https://www.jiscmail.ac.uk/policyandsecurity/

Re: [ccp4bb] Relion's own implementation of motioncorr2 on gpu cluster

[Paul Emsley]

On 11/09/2020 00:18, Lora Sinha wrote:
In the Relion 3.1 tutorial, the job is submitted to a local machine, 
but I wonder if I can submit the job to a GPU cluster? What MPI 
configuration should I use?



Is this the mailing list you're looking for?



To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1

This message was issued to members of www.jiscmail.ac.uk/CCP4BB, a mailing list 
hosted by www.jiscmail.ac.uk, terms & conditions are available at 
https://www.jiscmail.ac.uk/policyandsecurity/

[ccp4bb] Relion's own implementation of motioncorr2 on gpu cluster

[Lora Sinha]

In the Relion 3.1 tutorial, the job is submitted to a local machine, but I
wonder if I can submit the job to a GPU cluster? What MPI configuration
should I use?



To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1

This message was issued to members of www.jiscmail.ac.uk/CCP4BB, a mailing list 
hosted by www.jiscmail.ac.uk, terms & conditions are available at 
https://www.jiscmail.ac.uk/policyandsecurity/