AXI4MLIR: User-Driven Automatic Host Code Generation for Custom AXI-Based Accelerators
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2023
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866916169224552448 |
|---|---|
| author | Agostini, Nicolas Bohm Haris, Jude Gibson, Perry Jayaweera, Malith Rubin, Norm Tumeo, Antonino Abellán, José L. Cano, José Kaeli, David |
| author_facet | Agostini, Nicolas Bohm Haris, Jude Gibson, Perry Jayaweera, Malith Rubin, Norm Tumeo, Antonino Abellán, José L. Cano, José Kaeli, David |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2312_14821 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | AXI4MLIR: User-Driven Automatic Host Code Generation for Custom AXI-Based Accelerators Agostini, Nicolas Bohm Haris, Jude Gibson, Perry Jayaweera, Malith Rubin, Norm Tumeo, Antonino Abellán, José L. Cano, José Kaeli, David Programming Languages Hardware Architecture D.3.3 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. |
| title | AXI4MLIR: User-Driven Automatic Host Code Generation for Custom AXI-Based Accelerators |
| topic | Programming Languages Hardware Architecture D.3.3 |
| url | https://arxiv.org/abs/2312.14821 |