Make Every Move Count: LLM-based High-Quality RTL Code Generation Using MCTS
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866914667106926592 |
|---|---|
| author | DeLorenzo, Matthew Chowdhury, Animesh Basak Gohil, Vasudev Thakur, Shailja Karri, Ramesh Garg, Siddharth Rajendran, Jeyavijayan |
| author_facet | DeLorenzo, Matthew Chowdhury, Animesh Basak Gohil, Vasudev Thakur, Shailja Karri, Ramesh Garg, Siddharth Rajendran, Jeyavijayan |
| contents | Existing large language models (LLMs) for register transfer level code generation face challenges like compilation failures and suboptimal power, performance, and area (PPA) efficiency. This is due to the lack of PPA awareness in conventional transformer decoding algorithms. In response, we present an automated transformer decoding algorithm that integrates Monte Carlo tree-search for lookahead, guiding the transformer to produce compilable, functionally correct, and PPA-optimized code. Empirical evaluation with a fine-tuned language model on RTL codesets shows that our proposed technique consistently generates functionally correct code compared to prompting-only methods and effectively addresses the PPA-unawareness drawback of naive large language models. For the largest design generated by the state-of-the-art LLM (16-bit adder), our technique can achieve a 31.8% improvement in the area-delay product. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_03289 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Make Every Move Count: LLM-based High-Quality RTL Code Generation Using MCTS DeLorenzo, Matthew Chowdhury, Animesh Basak Gohil, Vasudev Thakur, Shailja Karri, Ramesh Garg, Siddharth Rajendran, Jeyavijayan Machine Learning Artificial Intelligence Hardware Architecture Existing large language models (LLMs) for register transfer level code generation face challenges like compilation failures and suboptimal power, performance, and area (PPA) efficiency. This is due to the lack of PPA awareness in conventional transformer decoding algorithms. In response, we present an automated transformer decoding algorithm that integrates Monte Carlo tree-search for lookahead, guiding the transformer to produce compilable, functionally correct, and PPA-optimized code. Empirical evaluation with a fine-tuned language model on RTL codesets shows that our proposed technique consistently generates functionally correct code compared to prompting-only methods and effectively addresses the PPA-unawareness drawback of naive large language models. For the largest design generated by the state-of-the-art LLM (16-bit adder), our technique can achieve a 31.8% improvement in the area-delay product. |
| title | Make Every Move Count: LLM-based High-Quality RTL Code Generation Using MCTS |
| topic | Machine Learning Artificial Intelligence Hardware Architecture |
| url | https://arxiv.org/abs/2402.03289 |