Compiler Support for Speculation in Decoupled Access/Execute Architectures
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866916579533389824 |
|---|---|
| author | Szafarczyk, Robert Nabi, Syed Waqar Vanderbauwhede, Wim |
| author_facet | Szafarczyk, Robert Nabi, Syed Waqar Vanderbauwhede, Wim |
| contents | Irregular codes are bottlenecked by memory and communication latency. Decoupled access/execute (DAE) is a common technique to tackle this problem. It relies on the compiler to separate memory address generation from the rest of the program, however, such a separation is not always possible due to control and data dependencies between the access and execute slices, resulting in a loss of decoupling.
In this paper, we present compiler support for speculation in DAE architectures that preserves decoupling in the face of control dependencies. We speculate memory requests in the access slice and poison mis-speculations in the execute slice without the need for replays or synchronization. Our transformation works on arbitrary, reducible control flow and is proven to preserve sequential consistency. We show that our approach applies to a wide range of architectural work on CPU/GPU prefetchers, CGRAs, and accelerators, enabling DAE on a wider range of codes than before. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_13553 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Compiler Support for Speculation in Decoupled Access/Execute Architectures Szafarczyk, Robert Nabi, Syed Waqar Vanderbauwhede, Wim Performance Hardware Architecture Distributed, Parallel, and Cluster Computing Irregular codes are bottlenecked by memory and communication latency. Decoupled access/execute (DAE) is a common technique to tackle this problem. It relies on the compiler to separate memory address generation from the rest of the program, however, such a separation is not always possible due to control and data dependencies between the access and execute slices, resulting in a loss of decoupling. In this paper, we present compiler support for speculation in DAE architectures that preserves decoupling in the face of control dependencies. We speculate memory requests in the access slice and poison mis-speculations in the execute slice without the need for replays or synchronization. Our transformation works on arbitrary, reducible control flow and is proven to preserve sequential consistency. We show that our approach applies to a wide range of architectural work on CPU/GPU prefetchers, CGRAs, and accelerators, enabling DAE on a wider range of codes than before. |
| title | Compiler Support for Speculation in Decoupled Access/Execute Architectures |
| topic | Performance Hardware Architecture Distributed, Parallel, and Cluster Computing |
| url | https://arxiv.org/abs/2501.13553 |