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CUDA_PROFILE_LOG文件中提供以下计数器:
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,有一个错误:
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CUDA 4.1。如何在命令行中获取DRAM_Reads和DRAM_Writes?
编辑:在做了一些读取后,我认为GPU可能有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
答案 0 :(得分:2)
由于硬件限制,并非所有计数器都可以在一次运行中进行分析。
根据警告消息,您可以尝试在第一次运行中分析前两个计数器,然后在第二次运行中分析最后两个计数器。