Guidance regarding output processing

[njuffa]

I wouldn’t recommend printing from device code except for debugging purposes. Why not simply copy the buffer from the device to the host first? You can apply whatever post-processing is desired or required after that.

FWIW, I recently posted a minimal example of printing from device code into a file here:

Can I print-to-file from a kernel?

There is only a device-side printf(), there is no device-side fprintf(). The way that device-side printf works is by depositing data into a buffer that is copied back to the host, and processed there via stdout. Note that the buffer can overflow if a kernel produces a lot of output. Programmers can select a size different from the default size (I seem to recall it is 1 MB) by specifying the desired size with a call to: cudaDeviceSetLimit (cudaLimitPrintfFifoSize, size_t size) On the host side,…