-
Notifications
You must be signed in to change notification settings - Fork 8
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[BBPBGLIB-712] Estimate memory usage for synapse and connection #22
Conversation
- Enabled Target methods to return raw gids - Accumulate counts in Counter() object (from collections) - MPI gather all counts, merge them and display
…pses - Added script used to measure synapse memory usage - Dry run now calculates estimate of memory usage of synapses - Fix typos
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
Nitpick: In the benchmark script we could work in Kilobytes to avoid multiplying by 1024 bellow. |
Thanks for the review @ferdonline . I think also we can integrate the json export into a different MR. |
Logfiles from GitLab pipeline #143299 (:white_check_mark:) have been uploaded here! Status and direct links: |
As detailed in [BBPBGLIB-712](https://bbpteam.epfl.ch/project/issues/browse/BBPBGLIB-712), this MR aims to introduce memory estimation for synapses and relative connections. A basic counter is introduced that tracks all the synapses that get instantiated on all the ranks. Then memory usage is estimated using a LUT containing the approximate memory estimate for each type of synapse. The values were calculated using the script `synstat.py` included in the `_benchmarks` folder. Based on the counter and the memory estimate, an approximated figure for memory utilization can be reported to the user. --------- Co-authored-by: Fernando Pereira <[email protected]>
As detailed in BBPBGLIB-712, this MR aims to introduce memory estimation for synapses and relative connections.
A basic counter is introduced that tracks all the synapses that get instantiated on all the ranks. Then memory usage is estimated using a LUT containing the approximate memory estimate for each type of synapse. The values were calculated using the script
synstat.py
included in the_benchmark
folder.Based on the counter and the memory estimate, an approximated figure for memory utilization can be reported to the user.