-
We are using a rayon thread pool to run a rust-based userspace driver. A mutex must be acquired before calls can be made to the driver.
We are running into a common deadlock problem with rayon wor…
-
Hello! I have an idea to reduce memory latency on NUMA or CCX-based systems.
I've read older issues here on this topic, where it is argued that work stealing is more important to performance than p…
-
We need a work stealing implementation (like the one Java MMTk has) to improve load balancing for the work-packets system.
-
nvidia's stdexec now has a work stealing queue:
https://github.com/NVIDIA/stdexec/commit/3a0439c9628a65edf3ed58e86d3a6b078a2d324b#diff-89920d7a167760061fc1c06555f9a0ac4955fb47d79769ca668c654d91ac60…
-
`st3` looks to be a fairly promising public crate for implementing `tokio`'s `!Send` local worker queues which eliminate most of the atomic operations when popping from the local queue.
Right now …
-
Thank you for writing and publishing this thread pool implementation. It seems like a glaring hole in the C++ STL / Boost etc. As it stands std::async is totally unusable (runs "deferred" or one threa…
-
I watched this video https://www.youtube.com/watch?v=gof_OEv71Aw.
Nicholas put up with an example explaining the term "steal work" mechanism of the crate "rayon", an array from 0 to 22. Other thre…
-
Hey, just thought you might want to know: it looks like the AI company [Cognition](https://www.cognition-labs.com/blog) is making it look like the code in this repo was written by their AI (named devi…
-
Currently each mapper executor fiber processes its file fully. This is simple and efficient approach but it creates sometimes a long tail latency when some mappers has finished but others are still p…
-
Right now I just used naive centralized task queue to manage task in the thread pool.
This may be not so efficient.
Try to write some benchmark for it and implement work stealing to improve the perfor…