VizWorld.com

Accelereyes has release Jacket GBENCH 1.0 for benchmarking GPU performance on your system across a wide variety of scientific algorithms, including LU Decomposition, FFT’s, BLAS, 3D Convolutions, and more.

GBENCH is a practical application benchmark measured in real seconds and is not meant to be a scientific or theoretical benchmark measured in GFLOPs. Also note that for fairness, arithmetic precisions (e.g. double, single) have been matched on the CPU and GPU. Finally, the data sizes used in these computations are large enough to exploit data parallelism (e.g. no scalar arithmetic was attempted). This benchmark assumes a data parallel problem.

It requires the MATLAB Compiler Runtime, but runs on both Windows and LInux.

via AccelerEyes – Jacket GBENCH – For GPU System Benchmarking.

Science accelereyes, benchmark, gbench, gpgpu

All of you Matlab folks out there might want to head on over to AccelerEyes and check out Jacket v1.1.  Big features in this release, including:

• Support for double-precision arithmetic. This enables a higher-level of accuracy for applications requiring fine precision.
• Expanded type support, including support for logical, int32, uint32, etc.
• New Developer SDK enables integration of custom CUDA kernels into the Jacket runtime. This allows custom code to inherit Jacket’s optimizations for memory transfers, kernel executions, and system performance within MATLAB.
• Expanded support for filtering functions, such as conv2, convn, filter2, etc, to include kernel sizes up to 10×10.
• New licensing scheme to support Concurrent Network licenses.
• Expanded support for reductions functions, such as sum, min, max, any, all, find.

Requires NVidia’s CUDA 2.2, and a 15-day trial is available to try before you buy.

20090622-AnnouncingJacket1.1.pdf (application/pdf Object).

Science accelereyes, cuda, gpgpu, jacket, matlab, nvidia