Low-latency hierarchical routing of reconfigurable neuromorphic systems
A reconfigurable hardware accelerator implementation for spiking neural network (SNN) simulation using field-programmable gate arrays (FPGAs) is promising and attractive research because massive parallelism results in better execution speed. For large-scale SNN simulations, a large number of FPGAs a...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-02-01
|
| Series: | Frontiers in Neuroscience |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fnins.2025.1493623/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832542285871448064 |
|---|---|
| author | Samalika Perera Ying Xu André van Schaik Runchun Wang |
| author_facet | Samalika Perera Ying Xu André van Schaik Runchun Wang |
| author_sort | Samalika Perera |
| collection | DOAJ |
| description | A reconfigurable hardware accelerator implementation for spiking neural network (SNN) simulation using field-programmable gate arrays (FPGAs) is promising and attractive research because massive parallelism results in better execution speed. For large-scale SNN simulations, a large number of FPGAs are needed. However, inter-FPGA communication bottlenecks cause congestion, data losses, and latency inefficiencies. In this work, we employed a hierarchical tree-based interconnection architecture for multi-FPGAs. This architecture is scalable as new branches can be added to a tree, maintaining a constant local bandwidth. The tree-based approach contrasts with linear Network on Chip (NoC), where congestion can arise from numerous connections. We propose a routing architecture that introduces an arbiter mechanism by employing stochastic arbitration considering data level queues of First In, First Out (FIFO) buffers. This mechanism effectively reduces the bottleneck caused by FIFO congestion, resulting in improved overall latency. Results present measurement data collected for performance analysis of latency. We compared the performance of the design using our proposed stochastic routing scheme to a traditional round-robin architecture. The results demonstrate that the stochastic arbiters achieve lower worst-case latency and improved overall performance compared to the round-robin arbiters. |
| format | Article |
| id | doaj-art-0e09a00a403342169380bcf534530d26 |
| institution | Kabale University |
| issn | 1662-453X |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Neuroscience |
| spelling | doaj-art-0e09a00a403342169380bcf534530d262025-02-04T06:32:02ZengFrontiers Media S.A.Frontiers in Neuroscience1662-453X2025-02-011910.3389/fnins.2025.14936231493623Low-latency hierarchical routing of reconfigurable neuromorphic systemsSamalika PereraYing XuAndré van SchaikRunchun WangA reconfigurable hardware accelerator implementation for spiking neural network (SNN) simulation using field-programmable gate arrays (FPGAs) is promising and attractive research because massive parallelism results in better execution speed. For large-scale SNN simulations, a large number of FPGAs are needed. However, inter-FPGA communication bottlenecks cause congestion, data losses, and latency inefficiencies. In this work, we employed a hierarchical tree-based interconnection architecture for multi-FPGAs. This architecture is scalable as new branches can be added to a tree, maintaining a constant local bandwidth. The tree-based approach contrasts with linear Network on Chip (NoC), where congestion can arise from numerous connections. We propose a routing architecture that introduces an arbiter mechanism by employing stochastic arbitration considering data level queues of First In, First Out (FIFO) buffers. This mechanism effectively reduces the bottleneck caused by FIFO congestion, resulting in improved overall latency. Results present measurement data collected for performance analysis of latency. We compared the performance of the design using our proposed stochastic routing scheme to a traditional round-robin architecture. The results demonstrate that the stochastic arbiters achieve lower worst-case latency and improved overall performance compared to the round-robin arbiters.https://www.frontiersin.org/articles/10.3389/fnins.2025.1493623/fullneuromorphic engineeringFPGA accelerationmulti-FPGAnetworks on chiptransceiversSNN |
| spellingShingle | Samalika Perera Ying Xu André van Schaik Runchun Wang Low-latency hierarchical routing of reconfigurable neuromorphic systems Frontiers in Neuroscience neuromorphic engineering FPGA acceleration multi-FPGA networks on chip transceivers SNN |
| title | Low-latency hierarchical routing of reconfigurable neuromorphic systems |
| title_full | Low-latency hierarchical routing of reconfigurable neuromorphic systems |
| title_fullStr | Low-latency hierarchical routing of reconfigurable neuromorphic systems |
| title_full_unstemmed | Low-latency hierarchical routing of reconfigurable neuromorphic systems |
| title_short | Low-latency hierarchical routing of reconfigurable neuromorphic systems |
| title_sort | low latency hierarchical routing of reconfigurable neuromorphic systems |
| topic | neuromorphic engineering FPGA acceleration multi-FPGA networks on chip transceivers SNN |
| url | https://www.frontiersin.org/articles/10.3389/fnins.2025.1493623/full |
| work_keys_str_mv | AT samalikaperera lowlatencyhierarchicalroutingofreconfigurableneuromorphicsystems AT yingxu lowlatencyhierarchicalroutingofreconfigurableneuromorphicsystems AT andrevanschaik lowlatencyhierarchicalroutingofreconfigurableneuromorphicsystems AT runchunwang lowlatencyhierarchicalroutingofreconfigurableneuromorphicsystems |