Performance

• Tuning CUDA with the GPU Memory Hierarchy · 2024-11-27

  Global, shared, and register memory each have distinct latency and bandwidth. Performance comes from the right access pattern.

• Computer Architecture: A Quantitative Approach (6th ed.)
• Computer Architecture: A Quantitative Approach (6th ed.)
• Computer Systems: A Programmer's Perspective (3rd ed.)
• Computer Systems: A Programmer's Perspective (3rd ed.)
• Improving the Scalability and Performance of a Rails Application: A Case Study with Consul