Windows-based parallel application development gets a boost

With Nvidia's Parallel Nsight Professional Edition now available free of charge, GPUs look even better for parallel apps

Nvidia has announced that Parallel Nsight Professional Edition is now available for all Visual Studio developers, free of charge. This is good news for the increasing number of developers creating highly parallel applications that use Nvidia CUDA-enabled GPUs.

As was discussed in the early days of CUDA, the high number of independent processors in a GPU are especially good for embarrassingly parallel problems, such as with 3-D rendering, face recognition, Monte Carlo simulation, particle physics event reconstruction, biological sequence searching, genetic algorithms, password guessing, and weather modeling -- any workload that can be divided into many parallel tasks. Since then, applications have appeared that use the GPU for speeding up video encoding, modeling fluid dynamics, doing electronic design automation, calculating financial options returns, ray tracing, oil and gas exploration, and medical imaging.

In the last year, GPU computing has quietly taken the high-performance computing world by storm, to the point where the fastest computer in the world, the Tianhe-1A unit in China at the National Supercomputing Center in Tianjin, uses 14,336 Intel Xeon CPUs and 7,168 Nvidia Tesla M2050 GPUs. However, the free GPU software development tools were difficult for programmers to use, and Parallel Nsight Professional Edition, which makes developing and debugging code to run on CPUs and GPUs almost as easy as developing for ordinary CPUs and integrates tightly with Microsoft Visual Studio, was relatively expensive, considering its special purpose.

According to Nvidia, it is providing the full Parallel Nsight Pro feature set at no cost to bring this tool into line with CUDA and its other development tools. This allows a broader range of developers -- who may not have the funding available to professionals -- to use high-level tools to unlock the speed of GPUs for their applications. Developers can download the latest version from Nvidia (registration required).

This article, "Windows-based parallel application development gets a boost," was originally published at Get the first word on what the important tech news really means with the InfoWorld Tech Watch blog. For the latest business technology news, follow on Twitter.