IThemal: Accurate, Portable and Fast Basic Block Throughput Estimation using Deep Neural Networks
© 2019 by the author(s). Predicting the number of clock cycles a processor takes to execute a block of assembly instructions in steady state (the throughput) is important for both compiler designers and performance engineers. Building an analytical model to do so is especially complicated in modern...
| Main Authors: | , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
2021
|
| Online Access: | https://hdl.handle.net/1721.1/137422 |
| _version_ | 1826204712603484160 |
|---|---|
| author | Mendis, C Renda, A Amarasinghe, S Carbin, M |
| author2 | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory |
| author_facet | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Mendis, C Renda, A Amarasinghe, S Carbin, M |
| author_sort | Mendis, C |
| collection | MIT |
| description | © 2019 by the author(s). Predicting the number of clock cycles a processor takes to execute a block of assembly instructions in steady state (the throughput) is important for both compiler designers and performance engineers. Building an analytical model to do so is especially complicated in modern x86-64 Complex Instruction Set Computer (CISC) machines with sophisticated processor microarchitectures in that it is tedious, error prone, and must be performed from scratch for each processor generation. In this paper we present Ithemal, the first tool which learns to predict the throughput of a set of instructions. Ithemal uses a hierarchical LSTM-based approach to predict throughput based on the opcodes and operands of instructions in a basic block. We show that Ithemal is more accurate than state-of-the-art hand-written tools currently used in compiler backends and static machine code analyzers. In particular, our model has less than half the error of state-of-the-art analytical models (LLVM's llvm-mca and Intel's IACA). Ithemal is also able to predict these throughput values just as fast as the aforementioned tools, and is easily ported across a variety of processor microarchitectures with minimal developer effort. |
| first_indexed | 2024-09-23T12:59:51Z |
| format | Article |
| id | mit-1721.1/137422 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T12:59:51Z |
| publishDate | 2021 |
| record_format | dspace |
| spelling | mit-1721.1/1374222023-02-14T19:04:16Z IThemal: Accurate, Portable and Fast Basic Block Throughput Estimation using Deep Neural Networks Mendis, C Renda, A Amarasinghe, S Carbin, M Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory © 2019 by the author(s). Predicting the number of clock cycles a processor takes to execute a block of assembly instructions in steady state (the throughput) is important for both compiler designers and performance engineers. Building an analytical model to do so is especially complicated in modern x86-64 Complex Instruction Set Computer (CISC) machines with sophisticated processor microarchitectures in that it is tedious, error prone, and must be performed from scratch for each processor generation. In this paper we present Ithemal, the first tool which learns to predict the throughput of a set of instructions. Ithemal uses a hierarchical LSTM-based approach to predict throughput based on the opcodes and operands of instructions in a basic block. We show that Ithemal is more accurate than state-of-the-art hand-written tools currently used in compiler backends and static machine code analyzers. In particular, our model has less than half the error of state-of-the-art analytical models (LLVM's llvm-mca and Intel's IACA). Ithemal is also able to predict these throughput values just as fast as the aforementioned tools, and is easily ported across a variety of processor microarchitectures with minimal developer effort. 2021-11-05T11:35:22Z 2021-11-05T11:35:22Z 2019-06 2020-11-23T20:16:17Z Article http://purl.org/eprint/type/ConferencePaper https://hdl.handle.net/1721.1/137422 Mendis, C, Renda, A, Amarasinghe, S and Carbin, M. 2019. "IThemal: Accurate, Portable and Fast Basic Block Throughput Estimation using Deep Neural Networks." 36th International Conference on Machine Learning, ICML 2019, 2019-June. en http://proceedings.mlr.press/v97/mendis19a/mendis19a.pdf 36th International Conference on Machine Learning, ICML 2019 Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf arXiv |
| spellingShingle | Mendis, C Renda, A Amarasinghe, S Carbin, M IThemal: Accurate, Portable and Fast Basic Block Throughput Estimation using Deep Neural Networks |
| title | IThemal: Accurate, Portable and Fast Basic Block Throughput Estimation using Deep Neural Networks |
| title_full | IThemal: Accurate, Portable and Fast Basic Block Throughput Estimation using Deep Neural Networks |
| title_fullStr | IThemal: Accurate, Portable and Fast Basic Block Throughput Estimation using Deep Neural Networks |
| title_full_unstemmed | IThemal: Accurate, Portable and Fast Basic Block Throughput Estimation using Deep Neural Networks |
| title_short | IThemal: Accurate, Portable and Fast Basic Block Throughput Estimation using Deep Neural Networks |
| title_sort | ithemal accurate portable and fast basic block throughput estimation using deep neural networks |
| url | https://hdl.handle.net/1721.1/137422 |
| work_keys_str_mv | AT mendisc ithemalaccurateportableandfastbasicblockthroughputestimationusingdeepneuralnetworks AT rendaa ithemalaccurateportableandfastbasicblockthroughputestimationusingdeepneuralnetworks AT amarasinghes ithemalaccurateportableandfastbasicblockthroughputestimationusingdeepneuralnetworks AT carbinm ithemalaccurateportableandfastbasicblockthroughputestimationusingdeepneuralnetworks |