Throughput-oriented kernel porting onto FPGAs

Keyphrases

• High-throughput 100%
• Programming Interface 100%
• In Tandem 50%
• Graph Coloring 50%
• Performance Efficiency 50%
• Code Motion 50%
• Heterogeneous Systems 50%
• Parallel Processing 50%
• Optimization Framework 50%
• Reconfigurable Devices 50%
• Efficient Mapping 50%
• CUDA Kernel 50%
• Custom Accelerators 50%
• Heterogeneous Architectures 50%
• Region Graph 50%
• Processing Advantage 50%
• Codes on Graphs 50%
• Code Optimization 50%
• Hierarchical Regions 50%
• Low Power Processing 50%
• Usability Requirements 50%

Computer Science

• High Throughput 100%
• Field Programmable Gate Arrays 100%
• Parallel Processing 50%
• Optimization Framework 50%
• Annotation 50%
• Graph Coloring 50%
• Reconfigurable Device 50%
• Heterogeneous System 50%
• Array Variable 50%
• Code Optimization 50%
• Heterogeneous Architecture 50%
• Usability Requirement 50%
• Graphics Processing Unit 50%