New Product Updates – Jacket v1.8, LibJacket v1.1

John MelonakosAnnouncements, CUDA Leave a Comment


Introduction Enhance your code with the fastest, most comprehensive library for GPU computing:

  • Jacket – the best GPU computing in MATLAB®.  Take a tour and compare!
  • LibJacket – the best way to kick start your CUDA development.  Take a tour!

Both products enable:

  • Manipulating vectors, matrices, and ND arrays
  • Support for single- and double-precision, boolean, real, and complex numbers
  • Hundreds of routines for arithmetic, linear algebra, statistics, imaging, signal processing, and more (full list: Jacket, LibJacket)
  • Thousands of lines of optimized code for any CUDA-capable GPU
New Product Features

  • Expanded support for the Signal Processing, Image Processing, and Statistics Libraries included with both Jacket and LibJacket
  • Faster linear algebra for special systems (e.g. symmetric, positive definite, triangular, etc.)
  • Enhanced visualizations
  • jacket_logoNew and updated examples: FDTD, Mandelbrot fractals, maximum-likelihood neural segmentation, MDS for genomics
  • Built with CUDA 4.0 for peak performance
Examples & Benchmarks

  • Recent case studies: video feature learning (Stanford), digital holography (REAL3d), hybrid LU decomposition (SAIC), GPU Computing seminar (U Aalborg), and more
  • Stanford benchmarks show Jacket outperforms the Parallel Computing Toolbox.
  • Benchmarks showing how LibJacket outperforms MKL, IPP, and other CPU libraries.
Quick links

* MATLAB® is a registered trademark of The MathWorks, as if you didn’t know that already 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *