GCNAX: A Flexible and Energy-efficient Accelerator for Graph Convolutional Neural Networks

Computer Science

• Energy Efficient 100%
• Graph Convolutional Network 100%
• Loop Fusion 18%
• Energy Efficiency 9%
• Data Access 9%
• Analysis Framework 9%
• Underlying Hardware 9%
• Energy Saving 9%
• Deep Learning 9%
• Resource Utilisation 9%
• Effective Approach 9%
• Varying Number 9%
• Redundant Data 9%
• Network Configuration 9%
• Data Analytics 9%

Keyphrases

• Energy-efficient Accelerator 100%
• Fusion Strategy 18%
• Process Graph 9%
• Execution Cycle 9%
• Rigid Loop 9%
• Deep Graph Learning 9%
• Buffer Structure 9%
• Novel Accelerators 9%
• Graph Data Analytics 9%
• Neural Network Configuration 9%
• Graph Datasets 9%