BLAS error in Potential Blocks - too many memory regions

Asked by weijie

Dear all,

I run the Yade program on a workstation, and the workstation configuration is Intel® Xeon(R) Gold 6130 CPU @ 2.10GHz × 64 with 64 cores and 128 threads. But when I run the program with Yadedaily -j128, the following problems occur:

BLAS: Program is Terminated. Because you tried to allocate too many memory regions.
Segmentation fault (core dumped)

I am running Yadedaily in Ubuntu 18.04 environment.I found that when I reduce the number of threads to 48 threads, the program can run normally.If I want to use all threads to run the program, what should I do?

Thanks in advance.
Jie

Question information

Language:
English Edit question
Status:
Expired
For:
Yade Edit question
Assignee:
No assignee Edit question
Last query:
Last reply:
Whiteboard:
The wildly inaccurate title was bothering me.
Revision history for this message
Jérôme Duriez (jduriez) said :
#1

Hi,

The error seems to come from an external library (blas). Telling some details about your simulation would have helped (use of FlowEngine ?) more than this very general title ;-)

And it looks like an inherent limitation of that library to not be able to run (at least the way you're running it) on so many cores.

If you really want to use all your CPU power here (under the assumption that it would be beneficial..), I guess you need to modify either YADE source code or blas library or both.

Revision history for this message
weijie (amandajoe) said :
#2

Hi,Jérôme

I encountered this problem in running an example file of PotentialBlocks[1].

Best regards,
Jie

[1]https://gitlab.com/yade-dev/trunk/-/blob/master/examples/PotentialBlocks/cubePBscaled.py

Revision history for this message
weijie (amandajoe) said :
#3

This problem occurs approximately when running to 1000 steps.

Revision history for this message
Robert Caulk (rcaulk) said :
#4

I'm sorry I had no choice but to improve the title of this question.

Revision history for this message
Launchpad Janitor (janitor) said :
#5

This question was expired because it remained in the 'Open' state without activity for the last 15 days.