NVidia Releases CUDA 5

NVidia announces the CUDA 5 production release, a new version of the GPU-friendly parallel computing platform that offers RDMA support and other enhancements.

According to the company, the new CUDA 5 platform makes development of GPU-accelerated applications faster and easier than ever. It includes support for dynamic parallelism, GPU-callable libraries, NVidia GPUDirect technology support for RDMA (remote direct memory access), and the NVidia Nsight Eclipse Edition integrated development environment (IDE).

CUDA 5 features dynamic parallelism, which means GPU threads can dynamically spawn new threads, allowing the GPU to adapt to the data. GPU-callable libraries allow developers to use dynamic parallelism for their own GPU-callable libraries. The “object linking” capability lets developers compile multiple CUDA source files into separate object files and link them into larger applications and libraries.

GPUDirect support for RDMA minimizes system memory bottlenecks, and the NVidia Nsight Eclipse Edition generates CUDA code quickly and easily. NVidia Nsight Eclipse Edition also lets programmers develop, debug, and profile GPU applications within the Eclipse-based IDE on Linux and Mac OS X platforms. An integrated analysis system provides automated performance analysis and step-by-step guidance to fix performance bottlenecks in the code.

Additionally, NVidia has launched an online resource center for CUDA programmers at: http://docs.nvidia.com. The site offers access to all CUDA developer documentation and technologies, including tools, code samples, libraries, APIs, and tuning and programming guides. CUDA 5 can be downloaded for free from the NVidia Developer Zone website.