Section: New Results

Progress threads placement for overlapping MPI non-blocking collectives using simultaneous multi-threading

Non-blocking collectives have been proposed so as to allow communications to be overlapped with computation in order to amortize the cost of MPI collective operations. To obtain a good overlap ratio, communications and computation have to run in parallel. To achieve this, different hardware and software techniques exists. Using dedicated cores for progress threads is one of them. However, some CPUs provide Simultaneous Multi-Threading, which is the ability for a core to have multiple hardware threads running simultaneously, sharing the same arithmetic units. We propose [18], [3] to use SMT to run progress threads to avoid dedicated cores allocation. We have run benchmarks on Haswell processors, using its Hyper-Threading capability, and get good results for both performance and overlap for inter-node communications. However, we have shown that Simultaneous Multi-Threading for intra-communications leads to bad performances due to contention on cache.