CIMR-V: An End-to-End SRAM-based CIM Accelerator with RISC-V for AI Edge Device

Computer Science

• Energy Efficiency 100%
• Edge AI 100%
• Accelerator Memory 100%
• Inference Model 66%
• Artificial Intelligence 66%
• Programmability 33%
• Memory Design 33%
• Memory Architecture 33%
• Type Instruction 33%
• Parallel Computer 33%
• Deep Learning Method 33%
• Feature Map 33%

Keyphrases

• RISC-V 100%
• In-memory 100%
• AI Edge Device 100%
• High Energy Efficiency 28%
• AI Models 28%
• Model Inference 28%
• Energy Efficiency 14%
• Latency 14%
• Proposed Design 14%
• Deep Learning 14%
• Programmability 14%
• Feature Map 14%
• Parallel Computing 14%
• Synergize 14%
• Highly Parallel 14%
• Limited Datasets 14%
• Long Latency 14%
• Loading Weight 14%
• Processing-In-Memory Architectures 14%
• Computing Efficiency 14%
• Data Movement 14%
• Max Pooling 14%
• Keyword Spotting 14%
• Layer Fusion 14%
• Weight Map 14%