Benchmarking MILC code with OpenMP and MPI

TL;DR

This paper evaluates the performance of MILC code using OpenMP and MPI on SMP architectures, highlighting the importance of hybrid parallelism for optimal resource utilization on modern supercomputers.

Contribution

It presents a detailed benchmarking study of MILC code with OpenMP and MPI, addressing the challenges of hybrid parallelism on SMP systems.

Findings

OpenMP and MPI hybrid parallelism improves performance on SMP architectures.

Challenges in converting MPI code to include OpenMP parallelism are discussed.

Benchmark results on IBM and Sun systems demonstrate the effectiveness of hybrid approaches.

Abstract

A trend in high performance computers that is becoming increasingly popular is the use of symmetric multiprocessing (SMP) rather than the older paradigm of MPP. MPI codes that ran and scaled well on MPP machines can often be run on an SMP machine using the vendor's version of MPI. However, this approach may not make optimal use of the (expensive) SMP hardware. More significantly, there are machines like Blue Horizon, an IBM SP with 8-way SMP nodes at the San Diego Supercomputer Center that can only support 4 MPI processes per node (with the current switch). On such a machine it is imperative to be able to use OpenMP parallelism on the node, and MPI between nodes. We describe the challenges of converting MILC MPI code to using a second level of OpenMP parallelism, and benchmarks on IBM and Sun computers.