What says (about temperature and hardware errors) following nvidia-smi report ?
Is the temperature value inappropriate high (and why in this case FAN speed is only 30%) ? Does the number of ECC errors says that there is serious hardware problems w/C2050 ?
nvidia-smi -a
==============NVSMI LOG==============
Timestamp : Sun Mar 20 01:49:21 2011
Driver Version : 260.19.26
GPU 0:
Product Name : Tesla C2050
PCI Device/Vendor ID : 6d110de
PCI Location ID : 0:2:0
Board Serial : 0321510020977
Display : Connected
Temperature : 72 C
Fan Speed : 30%
Utilization
GPU : 0%
Memory : 0%
Volatile ECC errors :
Single bit :
FB : 18
RF : 0
L1 : 0
L2 : 0
Total : 18
Double bit :
FB : 46596
RF : 0
L1 : 0
L2 : 0
Total : 46596
Aggregate ECC errors :
Single bit :
Total : 0
Double bit :
Total : 6
BTW, the System temp of mobo according lm-sensors is about 46 C, what looks OK.
Mikhail