Andrei Nesterov

Interestingly, the excessive awaiting points seems to be hurting performance. For me that is counter-intuitive, because I was thinking about them as zero-cost abstractions. There is almost 30% degradation in...

Marshall, thank you for the answer. That's nice to see the project is still under maintenance. We'll take another look and come with some suggestions on the plugin API if...

We hit the same problem on exactly the same versions (drivers 550.54.14 + toolkit 12.4), even though our application worked fine for earlier driver and toolkit versions. A solution was...

I can also confirm that it works on `0.8`.

> Can you try the devel images? If runtime images are necessary for you I can look into why they are different (I'm thinking the .12 at the end of...

> Can you try the devel images? I've just tried `nvidia/cuda:12.5.1-devel-ubuntu24.04`, it works fine. > Alternatively - You can disable dynamic loading in favor of using dynamic linking, and that...

I've upgraded to 0.11.8. In logs, I see only the following message. The same message as before. ```bash thread 'main' panicked at /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/cudarc-0.11.8/src/lib.rs:98:5: Unable to dynamically load the "cublas" shared...

I've got you. Just wanted to clarify that I on the latest version in case you want me to test something :-)

I've built `candle` with `cudarc` from the master branch. This issue has been resolved by [the latest commit](https://github.com/coreylowman/cudarc/commit/d7ac2b481cb637f7a73f2520ff2d12809285133f). Thanks. @coreylowman Are you planning on making a release?

[Fixed](https://github.com/coreylowman/cudarc/commit/d7ac2b481cb637f7a73f2520ff2d12809285133f)