Modeling the Potential of Message-Free Communication via CXL.mem

TL;DR

This paper introduces a performance evaluation toolchain and an extended model to predict the benefits of using CXL.mem for data exchange in HPC systems, focusing on MPI applications and enabling targeted optimizations.

Contribution

It presents a novel toolchain and performance model that analyze MPI data access patterns to identify potential speedups from CXL.mem integration.

Findings

The model accurately predicts potential performance gains from CXL.mem.

Validation on sample applications demonstrates the tool's effectiveness.

The approach enables targeted MPI optimizations for improved HPC performance.

Abstract

Heterogeneous memory technologies are increasingly important instruments in addressing the memory wall in HPC systems. While most are deployed in single node setups, CXL.mem is a technology that implements memories that can be attached to multiple nodes simultaneously, enabling shared memory pooling. This opens new possibilities, particularly for efficient inter-node communication. In this paper, we present a novel performance evaluation toolchain combined with an extended performance model for message-based communication, which can be used to predict potential performance benefits from using CXL.mem for data exchange. Our approach analyzes data access patterns of MPI applications: it analyzes on-node accesses to/from MPI buffers, as well as cross-node MPI traffic to gather a full understanding of the impact of memory performance. We combine this data in an extended performance model to predict which data transfers could benefit from direct CXL.mem implementations as compared to traditional MPI messages. Our model works on a per-MPI call granularity, allowing the identification and later optimizations of those MPI invocations in the code with the highest potential for speedup by using CXL.mem. For our toolchain, we extend the memory trace sampling tool Mitos and use it to extract data access behavior. In the post-processing step, the raw data is automatically analyzed to provide performance models for each individual MPI call. We validate the models on two sample applications -- a 2D heat transfer miniapp and the HPCG benchmark -- and use them to demonstrate their support for targeted optimizations by integrating CXL.mem.