A Multi-Bit Near-RRAM based Computing Macro with Highly Computing Parallelism for CNN Application

Keyphrases

• Accuracy Loss 20%
• Analog multiplication 20%
• Artificial Intelligence 20%
• Artificial Intelligence Models 20%
• Bit Weight 20%
• Capacitor-based 20%
• Constrained Edge 20%
• Convolution Operation 20%
• Data Transfer 20%
• Device Variations 20%
• Edge Devices 20%
• Emerging Approaches 20%
• Energy Efficiency 20%
• High Energy Efficiency 40%
• High-throughput 20%
• Highly Parallel 20%
• In-memory 60%
• Joint Optimization 20%
• Latency 20%
• Low Latency 40%
• Low Power Consumption 20%
• Memory Device 20%
• Memory Unit 20%
• Memory-based 40%
• Memory-centric Computing 100%
• Multi-bit 100%
• Multiplication Circuit 20%
• Multiply-accumulate 40%
• Multiply-accumulate Operation 20%
• Near-memory 40%
• Non-ideality 20%
• Operation Speed 20%
• Parallel Computing 20%
• Parallelization 100%
• Practical Implementation 20%
• Processing Unit 20%
• Resistive Random Access Memory (ReRAM) 100%
• Resource-constrained 20%
• Switched Capacitor 20%
• Wake Word 20%
• Weight Mapping 20%

Engineering

• Artificial Intelligence 40%
• Bit Input 20%
• Convolutional Neural Network 100%
• Data Transfer 20%
• Energy Conservation 20%
• Energy Efficiency 20%
• High Energy Efficiency 40%
• Joints (Structural Components) 20%
• Low Power Consumption 20%
• Output Bit 20%
• Parallel Computer 20%
• Parallelism 100%
• Processing Unit 20%
• Random Access Memory Device 20%
• Resistive Random Access Memory 100%
• Speed Operation 20%
• Switched Capacitor 20%

Computer Science

• Artificial Intelligence 20%
• Artificial Intelligence Model 20%
• Computer Hardware 20%
• Convolutional Neural Network 100%
• Energy Efficiency 60%
• High Throughput 20%
• Low Power Consumption 20%
• Memory Design 20%
• Parallel Computer 20%
• Parallel Computing 100%
• Processing Unit 20%
• Random Access Memory 100%