SOURCE: Rogue Wave Software

Rogue Wave Software

November 15, 2010 08:30 ET

Rogue Wave Software Releases Version 7.0 of the IMSL® Fortran Numerical Library With Additional Algorithms and Support for NVIDIA CUDA

NEW ORLEANS, LA--(Marketwire - November 15, 2010) - Today at SuperComputing 2010, Rogue Wave Software, Inc., a Battery Ventures portfolio company, announced the general availability of IMSL Fortran Numerical Library version 7.0. This release of the company's powerful math and statistics algorithms will allow users with supported hardware to simply link the IMSL Fortran Library with NVIDIA's CUDA BLAS to gain the significant performance improvement available by offloading processing to the GPU.

"We are pleased to see the IMSL Fortran Library enabling support for NVIDIA GPU acceleration," said Will Ramey, senior product manager for CUDA at NVIDIA. "This release provides developers with a broad suite of powerful algorithms that can leverage the leading GPU hardware, reducing their development time yet increasing performance of their applications."

"Our customers use our libraries to build sophisticated analytic capabilities into their applications. These applications don't have the luxury of languishing in poor performance, so it's natural that they are deployed on parallel architectures. Not only do the IMSL Libraries embody decades of numerical and statistical algorithm development, but they also shield developers from the need to be experts in parallel computation. With this release, developers can simply link their applications to this version of our libraries to automatically offload the CPU by performing linear algebra on the NVIDIA GPUs," said Dr. Ed Stewart, IMSL Product Manager, Rogue Wave Software. "This is another step in Rogue Wave's mission to make it easier to develop parallel, data-intensive applications."

In addition, there are many new and updated algorithms in this version of the IMSL Fortran Library including:

  • New Feynman-Kac algorithm that solves Black-Scholes problems
  • New time series modeling algorithms including Automatic ARIMA, Regression ARIMA, and AUTO_PARM to detect structural breaks in time series
  • Integration of ARPACK functions including Partial SVD and symmetric, non-symmetric and complex eigenvalue problems
  • New statistic functions including Partial Least Squares (PLS) Regression and Maximum Likelihood Estimation (MLE)
  • Additional statistical tools like random copula routines and several noncentral distributions

"We are very excited to bring version 7.0 of the IMSL Fortran Library to market," continued Dr. Stewart. "The inclusion of CUDA support, performance improvements and new analytic capabilities should benefit both our extensive user community and developers looking to embed advanced analytic capabilities in their applications."

Availability

IMSL Fortran Library 7.0 is available immediately. For more information on the IMSL Fortran Library, please visit: http://www.vni.com/products/imsl/fortran/. A white paper available for download includes architecture details and benchmark results for several routines: www.vni.com/company/whitepapers/NVIDIA_WPaper.pdf.

About Rogue Wave Software
Rogue Wave Software, Inc. is the largest independent provider of cross-platform software development tools and embedded components for the next generation of HPC applications. Rogue Wave marries High Performance Computing with High Productivity Computing to enable developers to harness the power of parallel applications and multicore computing. Rogue Wave products reduce the complexity of prototyping, developing, debugging, and optimizing multi-processor and data-intensive applications. Rogue Wave customers are industry leaders in the Global 2000, ISVs, OEMs, government laboratories and research institutions that leverage computationally-complex and data-intensive applications to enable innovation and outperform competitors. For more information, visit www.roguewave.com.

Contact Information

  • Media Contact:

    Sy Nayman
    Rogue Wave Software
    Phone: (303) 545-3197
    E-mail: Email Contact