The PAPI NVML component provides an interface to the nVidia Management Library (nvml).

The NVIDIA Management Library can be downloaded as part of the Tesla Development Kit.
<http://developer.nvidia.com/tesla-deployment-kit>

Before running the NVML component, the configure script for the NVML component 
must be executed in order to generate the Makefile which contains the 
configuration settings. This script needs to be executed only once. 

    % cd < papi_dir >/src/components/nvml
    % ./configure --with-nvml_incdir=< NVML INCLUDE PATH > --with-nvml_libdir=<NVML LIBRARY PATH > --with-cuda_incdir=<path> --with-cuda_libdir=<path>

Furthermore, the NVML component will be added to PAPI during the configuration of 
PAPI by adding the '--with-components=nvml' command line option to configure.
    % ./configure --with-components=nvml

At build-time the nVidia compiler, nvcc, needs to be in your path, as does the cuda run-time library (libcudart.so).


Please refer to http://developer.download.nvidia.com/assets/cuda/files/CUDADownloads/NVML/nvml.pdf 
for details about NVML library
