POSTER: An Optimized Predication Execution for SIMD Extensions
| Autor: | Juan M. Cebrian, Miquel Moreto, Adrian Barredo, Marc Casas, Mateo Valero |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
010302 applied physics
Out-of-order execution Data parallelism Computer science 0211 other engineering and technologies 021107 urban & regional planning 02 engineering and technology Parallel computing Energy consumption 01 natural sciences Vector processor 0103 physical sciences SIMD Divergence (statistics) Efficient energy use Euclidean vector |
| Zdroj: | PACT |
| DOI: | 10.1109/pact.2019.00054 |
| Popis: | Vector processing is a widely used technique to improve performance and energy efficiency in modern processors. Most of them rely on predication to support divergence control. However, performance and energy consumption in predicated instructions are usually independent on the number of true values in a mask. This means that the efficiency of the system becomes sub-optimal as vector length increases. In this work we propose the Optimized Predication Execution (OPE) technique. OPE delays the execution of sparse masked vector instructions sharing the same PC, extracts their active elements and creates a new dense instruction with a higher mask density. After executing such dense instruction, results are restored to the original sparse instructions. Our approach improves performance by up to 25% and reduces dynamic energy consumption by up to 43% on real applications with predication. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|