Almost all decent C compilers out there convert the following code into a single division/modulo operation…
int a, b, c, d;
b = c / d;
a = c % d;
Is this the same case for device functions as well?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Functions | 1 | 4660 | August 30, 2007 | |
| Compile entire compilation unit for device | 0 | 420 | April 11, 2018 | |
| Improving division in kernel How to use log2 in a kernel? | 1 | 6292 | May 18, 2009 | |
| Is it possible to divide one CUDA code into some parts of different CUDA codes? | 3 | 756 | July 23, 2013 | |
| Why would putting __device__ functions in the same file as kernels make them faster? | 4 | 721 | June 7, 2017 | |
| CUDA JIT compiler | 1 | 1815 | June 1, 2010 | |
| CUDA | 1 | 1131 | December 23, 2011 | |
| calling a __device__ functions inside kernels | 4 | 20505 | August 16, 2013 | |
| The way __device__ funcs get allocated | 3 | 1284 | October 22, 2009 | |
| Performances test | 4 | 1753 | May 21, 2012 |