AccelerEyes Jacket for MATLAB

Jacket Examples

Jacket includes extensive documentation and user examples to help developers in getting started with MATLAB GPU computing.

Examples include:

- Matrix Multiply Benchmarks
- FFT Benchmarks
- Convolution Benchmarks
- Neural Network Example
- Filtering Example
- Financial Example
- Graphics Toolbox (ocean surfaces & rain drops)

AccelerEyes Brochure

Jacket makes GPU computing ridiculously easy. Jacket enables scientists and engineers to continue programming in the familiar MATLAB environment while it transparently offloads computation onto the GPU. Developers simply mark data, via casting operations, which they want to be computed on the GPU. Beyond that, their MATLAB code can remain untouched.

Jacket is built around a runtime system which manages the execution of data parallel tasks. By leveraging the M-Language, Jacket achieves data parallelism efficiently with minimal changes to existing code bases. Jacket optimally configures kernels to run on SIMD hardware and automatically minimizes memory transfers between separate hardware devices. Jacket also takes advantage of hardware-specific memory hierarchies.

Finally, Jacket includes a system to manage memory and schedule computation for seamless integration of visualization and computation. Jacket now fully supports NVIDIA’s Tesla and CUDA products.

Trials are fully functional, enabling you
to explore Jacket prior to purchase

Graphics Toolbox

The Graphics Toolbox extends Jacket to seamlessly integrate computation with visualisation
making difficult to program, multi-threaded, and real time graphical displays effortless to achieve.

For example, by placing a single visualisation command at the end of a loop, data may be viewed as it is processed in-place on the GPU. Jacket automatically makes load-balancing decisions to optimally use GPU resources for compute as well as display. Further, the Graphics Toolbox exposes the entire OpenGL API and allows for interactive scene creation and rapid prototyping.