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

CPU and DRAM energy measurement #36

Closed
jaywonchung opened this issue Feb 22, 2024 · 4 comments
Closed

CPU and DRAM energy measurement #36

jaywonchung opened this issue Feb 22, 2024 · 4 comments
Labels
enhancement New feature or request

Comments

@jaywonchung
Copy link
Member

jaywonchung commented Feb 22, 2024

CPU and DRAM energy typically take only a small fraction of Deep Learning workloads and, since most of the heavy lifting computations are done by GPUs, it hasn't been that fruitful to study the optimization of their energy consumption.

However, for users who just want energy measurement, having CPU and DRAM measurement is just nice. So we should add it as an optional feature.

@jaywonchung jaywonchung added the enhancement New feature or request label Feb 22, 2024
@Sunt-ing
Copy link
Contributor

Any progress or thoughts on this? I want to measure the power consumption of the CPUs on my Ubuntu server.

@jaywonchung
Copy link
Member Author

I will prioritize this but currently we're low on bandwidth. Contributions are welcome, of course. In the meantime, for CPU power measurement, perhaps look into the Likwid powermeter, which I was planning to port into Zeus.

@Sunt-ing
Copy link
Contributor

Thanks for your reply and provided information!

@jaywonchung
Copy link
Member Author

CPU and DRAM energy measurement via Intel RAPL has now been implemented. We'll release v0.10.0 in about a week!

@Sunt-ing Please don't hesitate to let us know if something doesn't work well. Thanks!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
Projects
None yet
Development

No branches or pull requests

2 participants