Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Generic Spiking Architecture (GenSA)

Technical Report ·
DOI:https://doi.org/10.2172/1673456· OSTI ID:1673456
 [1];  [1]
  1. Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)

Neuromorphic devices are a rapidly growing area of interest in industry, with machines in production by IBM and Intel, among others. These devices promise to reduce size, weight and power (SWaP) costs while increasing resilience and facilitating high- performance computing (HPC). Each device will favor some set of algorithms, but this relationship has not been thoroughly studied. The field of neuromorphic computing is so new that existing devices were designed with merely estimated use-cases in mind. To better understand the fit between neuromorphic algorithms and machines, a simulated machine can be configured to any point in the design space. This will identify better choices of devices, and perhaps guide the market in new directions. The design of a generic spiking machine generalizes existing examples while also looking forward to devices that haven't been built yet. Each parameter is specified, along the approach/mechanism by which the relevant component is implemented in the simulator.

Research Organization:
Sandia National Laboratories (SNL-NM), Albuquerque, NM (United States)
Sponsoring Organization:
USDOE National Nuclear Security Administration (NNSA)
OSTI ID:
1673456
Report Number(s):
SAND--2020-10632; 691343
Country of Publication:
United States
Language:
English

Similar Records

Multi-Objective Optimization for Size and Resilience of Spiking Neural Networks
Conference · Tue Oct 01 00:00:00 EDT 2019 · OSTI ID:1606935

Resilience and Robustness of Spiking Neural Networks for Neuromorphic Systems
Conference · Wed Jul 01 00:00:00 EDT 2020 · OSTI ID:1671419

Related Subjects