Deploy on server to determine which public crates are the fastest for communicating in different architectures
main() to change which cores you'd like to be utilized before running.cargo run --releasepython3 analyze.py to view the results and the corresponding latency distributionsI'd like to split up the experiments to test mpmc / mpsc / spsc against each other in these different architectures. This will be included at some point but for now mpmc & mpsc have been favored as there are more use cases for them
async_broadcast) are tailored to specific features which incur some speed disadvantages. Thus, I recommend doing adequate research to determine the best crate for your architecture and only use this as a reference.main() function and change the parameters as needed. If you'd like more realistic non-jittery results, consider using isolcpus or the likepayload_size parameter was included in main() as well, feel free to change this at your discretion. The default is 0 bytesIf you'd like another crate added, have questions / ideas, please DM me on twitter @Dub0x3A or on discord 0xdub. Cheers and thanks
$ claude mcp add rust-channel-benchmark \
-- python -m otcore.mcp_server <graph>