I’d like to know if there is any Linux utility (or windows) for monitoring power-related readings ? for Fermi ?
Linux is better.
How can I read temperature and power (?) related readings at runtime ? if it is possible.
Are there separate event counters for that ? and various ways to see the readings ?
Thanks !
nvidia-smi maybe?
$ nvidia-smi
Wed Apr 4 16:52:24 2012
+------------------------------------------------------+
| NVIDIA-SMI 2.285.05 Driver Version: 285.05.32 |
|-------------------------------+----------------------+----------------------+
| Nb. Name | Bus Id Disp. | Volatile ECC SB / DB |
| Fan Temp Power Usage /Cap | Memory Usage | GPU Util. Compute M. |
|===============================+======================+======================|
| 0. Tesla C2050 | 0000:03:00.0 On | 0 0 |
| 30% 61 C P12 Off / Off | 0% 6MB / 2687MB | 0% Default |
|-------------------------------+----------------------+----------------------|
| 1. Tesla C2050 | 0000:04:00.0 Off | 0 0 |
| 30% 58 C P12 Off / Off | 0% 6MB / 2687MB | 0% Default |
|-------------------------------+----------------------+----------------------|
| Compute processes: GPU Memory |
| GPU PID Process name Usage |
|=============================================================================|
| No running compute processes found |
+-----------------------------------------------------------------------------+
Besides nvidia-smi you can also use NVML, which exposes management and monitoring functionality through an API.