Skip to content
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

Monitor - Add support for AMD GPU. #580

Merged
merged 10 commits into from
Nov 27, 2023
Merged

Monitor - Add support for AMD GPU. #580

merged 10 commits into from
Nov 27, 2023

Conversation

guoshzhao
Copy link
Contributor

Description
Add AMD support in monitor.

Major Revision

  • Add library pyrsmi to collect metrics.
  • Currently can get device_utilization, device_power, device_used_memory and device_total_memory.

@guoshzhao guoshzhao requested a review from a team as a code owner November 23, 2023 02:59
Copy link

codecov bot commented Nov 23, 2023

Codecov Report

Attention: 33 lines in your changes are missing coverage. Please review.

Comparison is base (1ad1c21) 86.77% compared to head (31c1be0) 86.41%.

Files Patch % Lines
superbench/common/utils/device_manager.py 54.16% 33 Missing ⚠️
Additional details and impacted files
@@            Coverage Diff             @@
##             main     #580      +/-   ##
==========================================
- Coverage   86.77%   86.41%   -0.37%     
==========================================
  Files          96       96              
  Lines        6475     6544      +69     
==========================================
+ Hits         5619     5655      +36     
- Misses        856      889      +33     
Flag Coverage Δ
cpu-python3.6-unit-test 71.76% <44.59%> (-0.09%) ⬇️
cpu-python3.7-unit-test 71.76% <44.59%> (-0.09%) ⬇️
cpu-python3.8-unit-test 72.20% <44.59%> (-0.10%) ⬇️
cuda-unit-test 84.34% <45.94%> (-0.44%) ⬇️
directx-unit-test 35.59% <47.29%> (+0.36%) ⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

@yukirora
Copy link
Contributor

will you change the exec function in executor.py, no way to enable monitor on rocm platform currently

if self.__get_platform() == Platform.CUDA:
                    monitor = Monitor(
                        None, int(self._sb_monitor_config.sample_duration or 10),
                        int(self._sb_monitor_config.sample_interval or 1), self.__get_monitor_path(benchmark_name)
                    )
                    monitor.start()
                else:
                    logger.warning('Monitor can not support ROCM/CPU platform.')

@cp5555 cp5555 mentioned this pull request Nov 23, 2023
30 tasks
@guoshzhao
Copy link
Contributor Author

will you change the exec function in executor.py, no way to enable monitor on rocm platform currently

if self.__get_platform() == Platform.CUDA:
                    monitor = Monitor(
                        None, int(self._sb_monitor_config.sample_duration or 10),
                        int(self._sb_monitor_config.sample_interval or 1), self.__get_monitor_path(benchmark_name)
                    )
                    monitor.start()
                else:
                    logger.warning('Monitor can not support ROCM/CPU platform.')

Good catch. Thanks.

@guoshzhao guoshzhao merged commit 028819b into main Nov 27, 2023
22 of 23 checks passed
@guoshzhao guoshzhao deleted the guzhao/amd_monitor branch November 27, 2023 10:45
@yukirora yukirora mentioned this pull request Dec 6, 2023
29 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants