IMPULSE: A 65-nm Digital Compute-in-Memory Macro with Fused Weights and Membrane Potential for Spike-Based Sequential Learning Tasks

Engineering

• Sparsity 100%
• Energy Conservation 66%
• Energy Efficiency 66%
• Energy Engineering 33%
• High Energy Efficiency 33%
• Recurrent Neural Network 33%

Computer Science

• Neural Network 100%
• Sparsity 60%
• Energy Efficiency 60%
• Memory Access 20%
• Recurrent Neural Network 20%
• Spatiotemporal Data 20%
• Computer Hardware 20%
• Data Mapping 20%

Neuroscience

• Membrane Potential 100%
• Nerve Cell Membrane 20%
• Recurrent Neural Network 20%