Magnetic Skyrmion-Based Spiking Neural Network for Pattern Recognition

Round 1

Reviewer 1 Report

In this article “Magnetic Skyrmion based Spiking Neural Network (SNN) for pattern recognition”, authors constructed SNN with novel magnetic skyrmion based leaky-integrate fire (LIF) spiking neuron and the skyrmionic synapse crossbar. The simulation results have proved the proposed idea better than presented in the scientific literature.

 Although simulation outcomes are promising, however, authors need to implement their proposed idea in a real industrial environment and validate their simulation results.

 

 Also, authors are strongly recommended to highlight their research novelty and contributions in a subsection in the Introduction section.

 

 Conclusion section needs to be improved with more details of their research outcomes and should add a separate paragraph for future work and recommendations.

 

 

Author Response

Please see the attachment for the responses to Reviewer 1's Comments.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper presents a new method to simulate neurons and their synaptic responses. The researchers propose an architecture which uses a deep neural network spiked by skyrmions. The method was evaluated using metric such as energy leakage, utilization and latency. The results showed that the proposed architecture had better energy efficiency and was better at lower switching voltages. The research has important applications in pattern recognition and serves as an alternative for traditional semiconductor based neural networks. The paper is well organized with sufficient details about the methodology. The diagrams help understand the circuit and data flow and proves beneficial for researchers interested in implementing the proposed architecture. As a result, the reviewer recommends an accept.

Author Response

Thank you very much for reading our paper carefully and giving the above positive comments. Thank you again and best regards.

Round 2

Reviewer 1 Report

The revised manuscript is significantly improved.