GPU Analysis Toolkit

June 2012, 373575A-01

The LabVIEW GPU Analysis Toolkit is a software package that allows you to write a LabVIEW application that offloads FFT or BLAS operations to a graphics processing unit (GPU). The toolkit is designed to offload operations with data sets large enough to exceed CPU capabilities or operations with significant amounts of parallel computations. To use the GPU Analysis Toolkit, you must have a GPU hardware device installed where it is accessible from the host processor.

Communication between LabVIEW and a GPU

The GPU Analysis Toolkit uses a software development kit called the LVGPU SDK. The LVGPU SDK creates a special execution environment, or context, in which to perform operations on a GPU. It communicates with a GPU through an API, which is responsible for managing GPU execution from the host. In this toolkit, the function wrappers for the FFT and BLAS operations already are built with the LVGPU SDK, and they specifically call the NVIDIA CUDA libraries and communicate with a GPU through an NVIDIA API. You can use the LVGPU SDK to build wrappers for implementing custom GPU functions to execute on any co-processor device as long as LabVIEW can call the external function.

To comment on National Instruments documentation, refer to the National Instruments website.

© 2012 National Instruments. All rights reserved.