Eventprop training for efficient neuromorphic applications This work was funded by the EPSRC, grants EP/V052241/1 and EP/S030964/1; and Intel Corporation. TS is also funded by a PhD studentship from the University of Sussex. Compute time was provided through Gauss Centre for Supercomputing applicati
Thomas Shoesmith 1 , James C. Knight 1 , Balazs Meszaros 1 , Jonathan Timcheck 2 , and Thomas Nowotny 1 · 2025
Read Eventprop training for efficient neuromorphic applications This work was funded by the EPSRC, grants EP/V052241/1 and EP/S030964/1; and Intel Corporation. TS is also funded by a PhD studentship from the University of Sussex. Compute time was provided through Gauss Centre for Supercomputing applicati by Thomas Shoesmith 1 , James C. Knight 1 , Balazs Meszaros 1 , Jonathan Timcheck 2 , and Thomas Nowotny 1 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-04341.txt.
Description
Shoesmith and four co-authors presented this work in twenty twenty-five on training brain-like networks for special chips. Neuromorphic chips copy the brain's event-based style and can run AI with far less energy. Training such spiking networks has been the hard part. The team trained networks on graphics cards with an event-based method, then moved them to Intel's Loihi 2 chip.
They tested the pipeline on keyword spotting, a task that listens for set words. Accuracy barely dropped when the network moved to the chip. Sorting one sample on Loihi 2 ran up to ten times faster than on a small edge computer. It also used about two hundred times less energy. The results support moving AI tasks from the cloud to local devices.
Catalog record
| CREATOR | Thomas Shoesmith 1 , James C. Knight 1 , Balazs Meszaros 1 , Jonathan Timcheck 2 , and Thomas Nowotny 1 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 24 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-04341