Hello, there. Recently I wrote a cuda program, when it runs on my PC, whose OS is Windows XP 32bit with GTX 275, there is no problem. But when it runs on another computer RHEL5.3 64bit with Tesla S1070, it seems the S1070 does not work, and the related results are ‘nan’ thing. Both of them use the CUDA3.0. So I am totally confused, does anyone have the same problem, and is there any difference between Windows and Linux.
Are you using the driver API?
Are you using the driver API?
No, I think not. So what is the difference between Runtime API and Driver API?
No, I think not. So what is the difference between Runtime API and Driver API?