12 August 2010
AccelerEyes has released version 1.4 of its Jacket software platform for Matlab, adding support the latest Nvidia GPUs based on the Fermi architecture (Tesla 20-series and GeForce GTX 4xx-series). Nvidia’s latest Fermi architecture offers 448 computational cores, increased IEEE-754 floating-point arithmetic precision, error-correcting memory for reliable computation, and enhanced memory-caching mechanisms.
Jacket is a software platform that delivers GPU computing power to desktop users of Matlab and other very high level languages. It enables faster prototyping and problem solving across a range of government, manufacturing, energy, media, biomedical, financial, and scientific research applications. The Jacket software platform enables accelerated double-precision performance for common arithmetic and linear algebra functionality on the new Nvidia hardware based on the Fermi architecture. Jacket eliminates the need to manually port Matlab code to C, C++, or CUDA to run on GPU-based workstations and clusters – a productivity step that can take months to years to complete for computationally intensive problems.
In addition to performance improvements provided by the Fermi hardware, AccelerEyes has improved the performance of Jacket's on-the-fly compilation technology. The performance of matrix multiplication functions has also been improved in the latest release, and the company claims 20-60 per cent speed-up for these tasks. Additional features have been introduced in Jacket 1.4, including direct access to the Nvidia performance primitives enabling new image processing functionality such as ERODE and DILATE.
Support for new Nvidia GPU platforms in Jacket 1.4 extends across Linux, Windows, and MacOS.