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.

Sign up for GitHub

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

Jump to bottom

Updates on How To Characterize Systems using GEOPM #3183

Open

lhlawson wants to merge 25 commits into geopm:dev from lhlawson:public-lhlawson-system-characterization-validation

Contributor

lhlawson commented Oct 5, 2023

Relates to System Characterization Tests #3175 from github issues
Fixes System Characterization Tests #3175 change request from github issues.

A PR for covering validation and documentation of system characterization using GEOPM

lhlawson force-pushed the public-lhlawson-system-characterization-validation branch 2 times, most recently from b25c601 to aa5de33 Compare

October 6, 2023 00:16

lhlawson mentioned this pull request

System Characterization Tests #3175

Open

15 tasks

avilcheslopez reviewed

View reviewed changes

service/docs/source/geopm_agent_cpu_activity.7.rst Show resolved Hide resolved

bgeltz requested changes

View reviewed changes

integration/test/test_gpu_activity_agent.py Outdated Show resolved Hide resolved

bgeltz requested changes

View reviewed changes

integration/test/test_gpu_activity_agent.py Outdated Show resolved Hide resolved

integration/test/test_gpu_activity_agent.py Outdated Show resolved Hide resolved

lhlawson force-pushed the public-lhlawson-system-characterization-validation branch 2 times, most recently from e820d4c to bc47c8a Compare

October 12, 2023 14:55

cmcantalupo force-pushed the dev branch from 2bd5c8d to 2cd441d Compare

October 18, 2023 16:32

lhlawson force-pushed the public-lhlawson-system-characterization-validation branch from 7da4305 to 1b46759 Compare

October 18, 2023 21:01

cmcantalupo force-pushed the dev branch from 2cd441d to 0859ce3 Compare

October 19, 2023 00:41

cmcantalupo force-pushed the dev branch from 54d8906 to 3a36fcd Compare

October 26, 2023 02:54

avilcheslopez force-pushed the public-lhlawson-system-characterization-validation branch from 0ef46b3 to af42e40 Compare

October 31, 2023 16:38

avilcheslopez reviewed

View reviewed changes

integration/test/test_gpu_activity_agent.py

+                                                     experiment_cli_args=experiment_cli_args)
+                      base_dir = 'test_gpu_activity_output'
+                      cls._gpu_ca_characterization = GPUCACharacterization(base_dir=base_dir)

Contributor

avilcheslopez Oct 31, 2023

Add comment explaining what's being done here.

integration/test/test_gpu_activity_agent.py Outdated

Comment on lines 68 to 81

-                          node_count=node_count,
-                          parres_cores_per_node=None,
-                          parres_gpus_per_node=None,
-                          parres_cores_per_rank=1,
-                          parres_init_setup=None,
-                          parres_exp_setup=None,
-                          parres_teardown=None,
-                          parres_args=None,
-                          trial_count=1,
-                          cool_off_time=3,
-                          enable_traces=False,
-                          enable_profile_traces=False,
-                          phi_list=None,

Contributor

avilcheslopez Oct 31, 2023

This should not have been removed.

cmcantalupo mentioned this pull request

Improve guide for running the integration tests so that it can be directly applied to Aurora including high level script #3233

Open

5 tasks

avilcheslopez force-pushed the public-lhlawson-system-characterization-validation branch from 1c0ae23 to e95b72c Compare

November 16, 2023 22:18

lhlawson and others added 14 commits

November 17, 2023 08:45


          Started documenation on How To Characterize Systems with GEOPM

aedb661

Signed-off-by: Lowren Lawson <lowren.h.lawson@intel.com>


          Adding documentation of GPU F-E characterization

2320d80

Signed-off-by: lhlawson <lowren.h.lawson@intel.com>


          Provided info on how to check GPU-CA Fe fallbacks

d3b8c8a

Signed-off-by: lhlawson <lowren.h.lawson@intel.com>


          Include documentation for PBS in CPU activity agent page

06b9967

Signed-off-by: Alejandro Vilches <alejandro.vilches@intel.com>


          Minor CPU CA agent doc improvement

4ea3cd5

Add comments to InitControl file describing each setting.

Signed-off-by: Alejandro Vilches <alejandro.vilches@intel.com>


          First pass gpu activity agent test multi-vendor support

0b3eea9

Signed-off-by: lhlawson <lowren.h.lawson@intel.com>


          Updated gpu-ca test to use get_service_config_value

4ff766f

Co-authored-by: Brad Geltz <brgeltz@gmail.com>


          Apply suggestions from code review: Remove unnecessary skip decorators

0a88b10

Co-authored-by: Brad Geltz <brgeltz@gmail.com>


          Added GPU Freq Sweep and Characterization to GPU-CA test

9ea652f

Signed-off-by: lhlawson <lowren.h.lawson@intel.com>


          WIP Use geopm-ctl-local with test_gpu_activity

11460a2

Signed-off-by: Brad Geltz <brad.geltz@intel.com>


          Fixups for test_cpu_characterization and test_gpu_activity

b018d99

Signed-off-by: Brad Geltz <brad.geltz@intel.com>

WIP Use geopm-ctl-local with test_gpu_activity

Signed-off-by: Brad Geltz <brad.geltz@intel.com>

Fix bad merge

Signed-off-by: lhlawson <lowren.h.lawson@intel.com>


          WIP add test_gpu_characterization test and logic to gpu-ca agent test

ff5e3b7

Signed-off-by: lowren.h.lawson@intel.com <lowren.h.lawson@intel.com>


          Add module for performing CPU CA characterization

9715a91

Signed-off-by: Alejandro Vilches <alejandro.vilches@intel.com>


          Refactor CPU CA tests to use CPU characterization module

6dfecd6

Signed-off-by: Alejandro Vilches <alejandro.vilches@intel.com>

avilcheslopez added 3 commits

November 17, 2023 08:45


          Add module for performing GPU CA characterization

2d50f02

Signed-off-by: Alejandro Vilches <alejandro.vilches@intel.com>


          Add GPU characterization integration tests

eb1ae90

Signed-off-by: Alejandro Vilches <alejandro.vilches@intel.com>


          Enable process mode for minife

61a0386

Also add minor improvements to CPUCharacterization to facilitate debug.

Signed-off-by: Alejandro Vilches <alejandro.vilches@intel.com>

avilcheslopez force-pushed the public-lhlawson-system-characterization-validation branch from e95b72c to 61a0386 Compare

November 17, 2023 16:45

lhlawson and others added 8 commits

November 17, 2023 15:37


          Fixed GPU Characterizatoin experiment config

14b85b9

Signed-off-by: lhlawson <lowren.h.lawson@intel.com>


          Get CPU-CA integration tests to pass

1b6e970

- Enable process mode and affinity for AIB and minife
- Don't LD_PRELOAD libgeopm
- Adjust problem size for CPU-CA characterization sweeps
- Misc. clean-up

Signed-off-by: Alejandro Vilches <alejandro.vilches@intel.com>


          Temp fixes for unblocking frequency sweep experiments on system

c028b7e

Signed-off-by: lhlawson <lowren.h.lawson@intel.com>


          Added report parsing updates for geopm-ctl-local runs

4a648cb

Signed-off-by: lhlawson <lowren.h.lawson@intel.com>


          Added preload and program filter to parres

0589e82

Signed-off-by: lhlawson <lowren.h.lawson@intel.com>


          Added handling for geopm-ctl-local named reports

1bf8b6f

Signed-off-by: lhlawson <lowren.h.lawson@intel.com>


          changed get_df to get_app_df to match recommendation script

a7a8c5b

Signed-off-by: lhlawson <lowren.h.lawson@intel.com>


          Update DGEMM GPU-CA test to remove < FOM check and increase iterations

ed4199f

Signed-off-by: lhlawson <lowren.h.lawson@intel.com>

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment