Dear all, I used 10 PyNQ nodes to run SNN using mpirun, which is faster than a single node. but I want to know how long each node takes to start MPI and wait for MPI_all_gether to end. Where do I need to print the time in the source code?Or what tools can I use to test MPI communication times?（my nest version is 2.14）
Best regards Cong
Dear Cong,
please have a look at https://nest-simulator.readthedocs.io/en/v3.3/guides/built-in_timers.html
Cheers, Jochen!
On July 5, 2022 3:56:09 AM GMT+02:00, 512491013@qq.com wrote:
Dear all, I used 10 PyNQ nodes to run SNN using mpirun, which is faster than a single node. but I want to know how long each node takes to start MPI and wait for MPI_all_gether to end. Where do I need to print the time in the source code?Or what tools can I use to test MPI communication times?（my nest version is 2.14）
Best regards Cong ________________________________ NEST Users mailing list -- users@nest-simulator.org To unsubscribe send an email to users-leave@nest-simulator.org
------------------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------------------ Forschungszentrum Juelich GmbH 52425 Juelich Sitz der Gesellschaft: Juelich Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 Vorsitzender des Aufsichtsrats: MinDir Volker Rieke Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender), Karsten Beneke (stellv. Vorsitzender), Prof. Dr. Astrid Lambrecht, Prof. Dr. Frauke Melchior ------------------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------------------
Neugierige sind herzlich willkommen am Sonntag, den 21. August 2022, von 10:00 bis 17:00 Uhr. Mehr unter: https://www.tagderneugier.de
Ah, yes: Some of these timers have been backported to NEST 2.14.1 https://github.com/nest/nest-simulator/releases/tag/v2.14.1, but you should definitely upgrade to the newest version unless there’s a really, really good reason to use one that’s 5 years old and contains hundreds of unfixed bugs and problems that have been addressed meanwhile…
Cheers, Jochen!
On 05.07.22 07:59, Jochen Martin Eppler wrote:
Dear Cong,
please have a look at https://nest-simulator.readthedocs.io/en/v3.3/guides/built-in_timers.html
Cheers, Jochen!
On July 5, 2022 3:56:09 AM GMT+02:00, 512491013@qq.com wrote:
Dear all, I used 10 PyNQ nodes to run SNN using mpirun, which is faster than a single node. but I want to know how long each node takes to start MPI and wait for MPI_all_gether to end. Where do I need to print the time in the source code?Or what tools can I use to test MPI communication times?（my nest version is 2.14） Best regards Cong ------------------------------------------------------------------------ NEST Users mailing list --users@nest-simulator.org To unsubscribe send an email tousers-leave@nest-simulator.org
Forschungszentrum Juelich GmbH 52425 Juelich Sitz der Gesellschaft: Juelich Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 Vorsitzender des Aufsichtsrats: MinDir Volker Rieke Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender), Karsten Beneke (stellv. Vorsitzender), Prof. Dr. Astrid Lambrecht, Prof. Dr. Frauke Melchior
Neugierige sind herzlich willkommen am Sonntag, den 21. August 2022, von 10:00 bis 17:00 Uhr. Mehr unter: https://www.tagderneugier.de
_______________________________________________NEST Users mailing list --users@nest-simulator.orgTo unsubscribe send an email tousers-leave@nest-simulator.org
Dear,Jochen Martin Eppler
My predecessors used 2.14 for some research, so it has been used until now, and now it seems to be time for an update, with a lot of useful features in 3.3. I also want to ask a question, the time I want to obtain MPI is to verify the performance of neuron redistribution algorithm. According to my knowledge of the source code, nest's neuron distribution algorithm is distributed through traversal VP. If I want to specify which process neurons are distributed to, can I only modify the traversal part in the source code? Or does 3.3 have a way to easily redistribute neurons?
Best wishes, Cong
-
512491013＠qq.com -
Jochen Martin Eppler