CUDA 10 Memory Transaction的一个现象

1. Introduction近日,在写一些microbenchmark分析cuda程序访存问题时,发现了一个有趣的问题。目前尚未找到合理的解释,先记录下来以待后续分析。实验平台为:NVIDIA GTX950,sm5.0,maxwell架构。2. Global MemoryA memory "request" is an instruction which accesses memory, and a "transaction" is the movement of a unit of da more ...

CUDA Sanitizer Samples使用

1. IntroductionCUDA 10.1推出了新的API:The Compute Sanitizer API,提供了更底层更丰富的Instrumentation API。https://docs.nvidia.com/cuda/sanitizer-docs/SanitizerApiGuide/index.html目前相关文档还比较简单,本文记录下官方Samp more ...