Tutorial: Introduction to CUDA Fortran | GTC 2013
Tutorial: Introduction to CUDA Fortran | GTC 2013
Tutorial: Introduction to CUDA Fortran | GTC 2013
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Compilation<br />
• pgfortran - PGI’s <strong>Fortran</strong> compiler<br />
– All source code with .cuf or .CUF is compiled as <strong>CUDA</strong> <strong>Fortran</strong> enabled<br />
au<strong>to</strong>matically<br />
– Flag <strong>to</strong> target architecture (eg. -Mcuda=cc35)<br />
– -Mcuda=emu specifies emulation mode<br />
– Flag <strong>to</strong> target <strong>CUDA</strong> Toolkit version (eg. -Mcuda=cuda5.0)<br />
– -Mcuda=fastmath enables faster intrinsics (__sinf())<br />
– -Mcuda=nofma turns off fused multiply-add<br />
– -Mcuda=maxregcount: limits register usage per thread<br />
– -Mcuda=ptxinfo prints memory usage per kernel<br />
• pgfortran -Mcuda -help for a full list