HYDRA: Hybrid Data Multiplexing and Run-time Layer Configurable DNN Accelerator
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910079983288320 |
|---|---|
| author | Kumar, Sonu Gupta, Komal Raut, Gopal Lokhande, Mukul Vishvakarma, Santosh Kumar |
| author_facet | Kumar, Sonu Gupta, Komal Raut, Gopal Lokhande, Mukul Vishvakarma, Santosh Kumar |
| contents | Deep neural networks (DNNs) offer plenty of challenges in executing efficient computation at edge nodes, primarily due to the huge hardware resource demands. The article proposes HYDRA, hybrid data multiplexing, and runtime layer configurable DNN accelerators to overcome the drawbacks. The work proposes a layer-multiplexed approach, which further reuses a single activation function within the execution of a single layer with improved Fused-Multiply-Accumulate (FMA). The proposed approach works in iterative mode to reuse the same hardware and execute different layers in a configurable fashion. The proposed architectures achieve reductions over 90% of power consumption and resource utilization improvements of state-of-the-art works, with 35.21 TOPSW. The proposed architecture reduces the area overhead (N-1) times required in bandwidth, AF and layer architecture. This work shows HYDRA architecture supports optimal DNN computations while improving performance on resource-constrained edge devices. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_04976 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | HYDRA: Hybrid Data Multiplexing and Run-time Layer Configurable DNN Accelerator Kumar, Sonu Gupta, Komal Raut, Gopal Lokhande, Mukul Vishvakarma, Santosh Kumar Hardware Architecture Artificial Intelligence Computer Vision and Pattern Recognition Image and Video Processing Deep neural networks (DNNs) offer plenty of challenges in executing efficient computation at edge nodes, primarily due to the huge hardware resource demands. The article proposes HYDRA, hybrid data multiplexing, and runtime layer configurable DNN accelerators to overcome the drawbacks. The work proposes a layer-multiplexed approach, which further reuses a single activation function within the execution of a single layer with improved Fused-Multiply-Accumulate (FMA). The proposed approach works in iterative mode to reuse the same hardware and execute different layers in a configurable fashion. The proposed architectures achieve reductions over 90% of power consumption and resource utilization improvements of state-of-the-art works, with 35.21 TOPSW. The proposed architecture reduces the area overhead (N-1) times required in bandwidth, AF and layer architecture. This work shows HYDRA architecture supports optimal DNN computations while improving performance on resource-constrained edge devices. |
| title | HYDRA: Hybrid Data Multiplexing and Run-time Layer Configurable DNN Accelerator |
| topic | Hardware Architecture Artificial Intelligence Computer Vision and Pattern Recognition Image and Video Processing |
| url | https://arxiv.org/abs/2409.04976 |