da4ml: Distributed Arithmetic for Real-time Neural Networks on FPGAs
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_ | 1866915953548197888 |
|---|---|
| author | Sun, Chang Que, Zhiqiang Loncar, Vladimir Luk, Wayne Spiropulu, Maria |
| author_facet | Sun, Chang Que, Zhiqiang Loncar, Vladimir Luk, Wayne Spiropulu, Maria |
| contents | Neural networks with a latency requirement on the order of microseconds, like the ones used at the CERN Large Hadron Collider, are typically deployed on FPGAs fully unrolled and pipelined. A bottleneck for the deployment of such neural networks is area utilization, which is directly related to the required constant matrix-vector multiplication (CMVM) operations. In this work, we propose an efficient algorithm for implementing CMVM operations with distributed arithmetic on FPGAs that simultaneously optimizes for area consumption and latency. The algorithm achieves resource reduction similar to state-of-the-art algorithms while being significantly faster to compute. The proposed algorithm is open-sourced and integrated into the \texttt{hls4ml} library, a free and open-source library for running real-time neural network inference on FPGAs. We show that the proposed algorithm can reduce on-chip resources by up to a third for realistic, highly quantized neural networks while simultaneously reducing latency, enabling the implementation of previously infeasible networks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_04535 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | da4ml: Distributed Arithmetic for Real-time Neural Networks on FPGAs Sun, Chang Que, Zhiqiang Loncar, Vladimir Luk, Wayne Spiropulu, Maria Hardware Architecture Machine Learning High Energy Physics - Experiment B.2.4; B.6 Neural networks with a latency requirement on the order of microseconds, like the ones used at the CERN Large Hadron Collider, are typically deployed on FPGAs fully unrolled and pipelined. A bottleneck for the deployment of such neural networks is area utilization, which is directly related to the required constant matrix-vector multiplication (CMVM) operations. In this work, we propose an efficient algorithm for implementing CMVM operations with distributed arithmetic on FPGAs that simultaneously optimizes for area consumption and latency. The algorithm achieves resource reduction similar to state-of-the-art algorithms while being significantly faster to compute. The proposed algorithm is open-sourced and integrated into the \texttt{hls4ml} library, a free and open-source library for running real-time neural network inference on FPGAs. We show that the proposed algorithm can reduce on-chip resources by up to a third for realistic, highly quantized neural networks while simultaneously reducing latency, enabling the implementation of previously infeasible networks. |
| title | da4ml: Distributed Arithmetic for Real-time Neural Networks on FPGAs |
| topic | Hardware Architecture Machine Learning High Energy Physics - Experiment B.2.4; B.6 |
| url | https://arxiv.org/abs/2507.04535 |