Magnetic domain wall based synaptic and activation function generator for neuromorphic accelerators

• Main Authors: Siddiqui, Saima Afroz (Author), Dutta, Sumit (Author), Tang, Astera S. (Author), Liu, Luqiao (Author), Ross, Caroline A. (Author), Baldo, Marc A (Author)
• Other Authors: Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science (Contributor), Massachusetts Institute of Technology. Department of Materials Science and Engineering (Contributor)
• Format: Article
• Language: English
• Published: American Chemical Society (ACS), 2020-09-30T14:30:22Z.
• Subjects: Article
• Online Access: Get fulltext

Magnetic domain walls are information tokens in both logic and memory devices and hold particular interest in applications such as neuromorphic accelerators that combine logic in memory. Here, we show that devices based on the electrical manipulation of magnetic domain walls are capable of implementing linear, as well as programmable nonlinear, functions. Unlike other approaches, domain-wall-based devices are ideal for application to both synaptic weight generators and thresholding in deep neural networks. Prototype micrometer-size devices operate with 8 ns current pulses and the energy consumption required for weight modulation is ≤16 pJ. Both speed and energy consumption compare favorably to other synaptic nonvolatile devices, with the expected energy dissipation for scaled 20 nm devices close to that of biological neurons.
National Science Foundation (U.S.) (Award 1639921)