I use Ray <https://github.com/ray-project/ray>, a unified framework for scaling AI and Python applications, in an HPC cluster. Ray is based on gRPC. The cluster I used has InfiniBand which has low latency and high bandwidth. I try to use IPoIB(Internet Protocol over InfiniBand). But in this way, I can not make full use of IB's bandwidth. I want to optimize Ray on an HPC cluster with InfiniBand. RDMA, which can reduce CPU interruptions for network processing and increase CPU utilization, allows for a better performance of the software. I think it is a good choice for me. But so far, gRPC does not support RDMA. May I ask why not? What should I do if I want to take advantage of the RMDA and IB networks?
-- You received this message because you are subscribed to the Google Groups "grpc.io" group. To unsubscribe from this group and stop receiving emails from it, send an email to grpc-io+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/grpc-io/960f4a61-577b-4151-8be2-3772d8ef13a0n%40googlegroups.com.
