I heard that better to disable SLI mode but after it, cudaGetDeviceCount returns 1 (not 2). Also I can’t use cudaSetDevice(1) - “invalid argument”. GPU-Z recognizes two GPUs but HWMonitor sees only one. I’m using mobile version of GF8s. Is here sth wrong with code or device?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| 2 Graphic cards, cudaGetDeviceCount() find only 1 | 4 | 1311 | September 16, 2009 | |
| 3 GTX 260 cards cudaGetDeviceCount | 0 | 1876 | June 24, 2009 | |
| cudaGetDeviceCount() only detects 1 GPU | 11 | 5444 | November 7, 2008 | |
| CUDA - Number of Devices Available Fewer devices available when SLI is disabled. Slower with SLI ena | 3 | 4065 | October 21, 2009 | |
| How to install 2 8800 GTX cards? | 0 | 4076 | April 11, 2007 | |
| CUDA 11.2 cudaGetDeviceCount bug | 1 | 592 | December 19, 2020 | |
| cudaGetDeviceCount returned 999 | 1 | 1818 | December 5, 2021 | |
| cudaGetDeviceCount() returns 1 for M60 Gpu? Should'n it be 2? | 6 | 1007 | November 7, 2017 | |
| cutil cuda sdk utility | 5 | 4981 | January 2, 2008 | |
| How to check if GPU supports CUDA? | 4 | 15750 | July 27, 2011 |