Rilasciata da NVidia la CUDA C Programming Best Practices Guide. La guida sarà di aiuto agli sviluppatori C nell’uso di CUDA.

This guide is designed to help developers programming for the CUDA architecture using C with CUDA extensions implement high performance parallel algorithms and understand best practices for GPU Computing. Chapters on the following topics and more are included in the guide:

• Introduction to Parallel Computing with CUDA
• Performance Metrics
• Memory Optimizations
• Execution Configuration Optimizations
• Instruction Optimizations
• Control Flow
• Debugging
• Numerical Accuracy and Precision
• Performance Optimization Strategies

This will be included with the 2.3 toolkit, but we decided to release it now because it’s definitely worthwhile reading for any CUDA C developer (a lot of collected internal wisdom on proven optimization strategies, for example).

La guida, in formato pdf, può essere scaricata da questo indirizzo: CUDA C Programming Best Practices Guide.

• 

Pubblicato nella Categoria: C/C++, Programmazione.
Tags: CUDA, nvidia.
Commenti