eprop.pdf (438.18 kB)
Efficient GPU training of LSNNs using eProp
conference contribution
posted on 2023-06-10, 05:18 authored by James KnightJames Knight, Thomas NowotnyThomas NowotnyTaking inspiration from machine learning libraries - where techniques such as parallel batch training minimise latency and maximise GPU occupancy - as well as our previous research on efficiently simulating Spiking Neural Networks (SNNs) on GPUs for computational neuroscience, we have extended our GeNN SNN simulator to enable spike-based machine learning research on general purpose hardware. We demonstrate that SNN classifiers implemented using GeNN and trained using the eProp learning rule can provide comparable performance to those trained using Back Propagation Through Time and show that the latency and energy usage of our SNN classifiers is up to 7 × lower than an LSTM running on the same GPU hardware.
History
Publication status
- Published
File Version
- Accepted version
Journal
ACM International Conference Proceeding SeriesPublisher
ACMExternal DOI
Page range
8-10Event name
NICE 2022: Neuro-Inspired Computational Elements ConferenceEvent location
Virtual Event, USAEvent type
conferenceEvent date
28th March - 1st AprilPlace of publication
New York, USAISBN
9781450395595Department affiliated with
- Informatics Publications
Full text available
- Yes
Peer reviewed?
- Yes
Legacy Posted Date
2022-11-02First Open Access (FOA) Date
2022-11-02First Compliant Deposit (FCD) Date
2022-11-02Usage metrics
Categories
No categories selectedKeywords
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC