NVML API Reference

1.0

The NVIDIA Management Library (NVML) is a C-based programatic interface for monitoring and managing various states within NVIDIA Tesla GPUs.
It is intended to be a platform for building 3rd party applications, and is also the underlying library for the NVIDIA-supported nvidia-smi tool.
NVML is thread-safe.


API Documentation

Supported OS platforms:

Supported products:

Support Methods:

Query Methods:

Control Methods:


Device Feature Matrix

This chart shows which features are available for each GPU product.
An updated version of the board's inforom may be required for some features.

Devices Tesla Architecture Fermi Architecture
Tesla
C1060
Tesla
S1070
Tesla
C2050
Tesla
C2070
Tesla
M2050
Tesla
M2070-Q
Tesla
M2070
Tesla
S2050
Tesla
X2070
display mode
persistence mode (linux-only)
driver model (win7-only)
serial
uuid
pci info
fan speed
compute mode
inforom version oem object
ecc object
pwr object
memory usage total
used
free
temperature gpu temp
mem temp
board temp
vr temp
power readings power state
power capping
power draw
clocks core clock
sm clock
mem clock
utilization gpu util
memory util
ecc mode current ecc
future ecc
volatile sbit
ecc errors
by location
total
volatile dbit
ecc errors
by location
total
aggregate sbit
ecc errors
by location
total
aggregate dbit
ecc errors
by location
total



Unit Feature Matrix

This chart shows which unit-level features are available for each S-class product.
All GPUs within each S-class product also provide the information listed in the Device chart above.

Units Tesla Architecture Fermi Architecture
Tesla
S1070
Tesla
S2050
id
serial
firmware
attached gpus
led state color
cause
temperature intake
exhaust
board
psu psu state
voltage
current
fan info fan speed
fan state


Generated on Tue Feb 8 09:10:20 2011