AXI4MLIR: User-Driven Automatic Host Code Generation for Custom AXI-Based Accelerators

December 22, 2023 · Entered Twilight · 🏛 IEEE/ACM International Symposium on Code Generation and Optimization

Repo contents: .clang-format, .clang-tidy, .dockerignore, .gitignore, .gitmodules, README.md, api_tester, benchmark, benchmark_conv, build-docker.sh, build_tools, cross-comp, docker, hw_parser, llvm-project, mlir-examples, potential_extensions, sample_code, scripts, start-docker.sh, tinybert

Authors Nicolas Bohm Agostini, Jude Haris, Perry Gibson, Malith Jayaweera, Norm Rubin, Antonino Tumeo, José L. Abellán, José Cano, David Kaeli arXiv ID 2312.14821 Category cs.PL: Programming Languages Cross-listed cs.AR Citations 3 Venue IEEE/ACM International Symposium on Code Generation and Optimization Repository https://github.com/AXI4MLIR/axi4mlir ⭐ 8 Last Checked 3 months ago

Abstract

This paper addresses the need for automatic and efficient generation of host driver code for arbitrary custom AXI-based accelerators targeting linear algebra algorithms, an important workload in various applications, including machine learning and scientific computing. While existing tools have focused on automating accelerator prototyping, little attention has been paid to the host-accelerator interaction. This paper introduces AXI4MLIR, an extension of the MLIR compiler framework designed to facilitate the automated generation of host-accelerator driver code. With new MLIR attributes and transformations, AXI4MLIR empowers users to specify accelerator features (including their instructions) and communication patterns and exploit the host memory hierarchy. We demonstrate AXI4MLIR's versatility across different types of accelerators and problems, showcasing significant CPU cache reference reductions (up to 56%) and up to a 1.65x speedup compared to manually optimized driver code implementations. AXI4MLIR implementation is open-source and available at: https://github.com/AXI4MLIR/axi4mlir.