0

我正在尝试在命令行中使用 CUDA Profiler；我对 DRAM_Reads 和 DRAM_Writes 感兴趣 - 我在我的 CUDA_PROFILE_LOG 文件中提供了以下计数器：

fb_subp0_read_sectors
fb_subp0_write_sectors
fb0_subp0_read_sectors
fb0_subp0_write_sectors
fb1_subp0_read_sectors
fb1_subp0_write_sectors

但我注意到在我的 cuda_profile 文件中，有如下错误：

NV_Warning: Ignoring the invalid profiler config option: fb0_subp0_read_sectors
NV_Warning: Ignoring the invalid profiler config option: fb0_subp0_write_sectors
NV_Warning: Ignoring the invalid profiler config option: fb1_subp0_read_sectors
NV_Warning: Ignoring the invalid profiler config option: fb1_subp0_write_sectors

我从 fb_subp0_read_sectors 和 fb_subp0_write_sectors 计数器获得的值不等于我从 NVidia Visual Profiler 获得的值，这可能是因为我没有将正确的计数器传递给配置文件。GPU是Tesla M2050，使用的是CUDA 4.1。如何在命令行中获取 DRAM_Reads 和 DRAM_Writes？

编辑：在做了一些阅读之后，我认为 GPU 可以有 fb0/1... 或 fb... 计数器。但即使我有：

fb_subp0_read_sectors
fb_subp0_write_sectors
fb_subp1_read_sectors
fb_subp1_write_sectors

我收到警告：

NV_Warning: Counter 'fb_subp1_read_sectors' is not compatible with other selected counters and it cannot be profiled in this run.
NV_Warning: Counter 'fb_subp1_write_sectors' is not compatible with other selected counters and it cannot be profiled in this run.

谢谢，萨扬

4

1 回答 1

2

Not all counters can be profiled in a single run, due to hardware constraints.

According to the warning message, you may try profiling the first two counters in the first run, and then the last two in the second run.

于 2012-07-11T03:49:28.593 回答