Software integrates desktop tools into HPC environment.
Share:
Press Release Summary:
Serving as open software platform, Star-P(TM) v2.4 technical computing software enables scientists and engineers to develop algorithms and models on PCs using desktop tools and run them interactively on parallel servers and clusters. In addition to MATLAB®, software supports open source Python programming language and Mathematica desktop environment. Star-P OpenConnect API lets users plug in to numerical libraries and solvers.
Original Press Release:
Star-P Opens Up
New Open Platform Architecture Supports a Wide Range of Third-Party Desktop Applications, Numerical Libraries and Hardware Accelerators
TAMPA, Fla., Nov. 13 -- Interactive Supercomputing (ISC) today unveiled at the SC06 show a new version of its Star-P(TM) technical computing software that makes it easy to integrate software and hardware from a variety of high performance computing (HPC) sources. Star-P 2.4 serves as an open software platform that supports many popular desktop tools, numerical libraries and hardware accelerators.
Star-P enables scientists, engineers and analysts to develop algorithms and models on PCs using familiar desktop tools, and run them instantly and interactively on parallel servers and clusters without having to re-program the code.
Supports more desktop apps
The new version of Star-P will add support for the open source Python programming language and Wolfram Research's Mathematica desktop environment. Combined with its existing support for MATLAB®, Star-P will be available to millions of users of the world's most popular mathematical desktop applications.
"We are excited to see the new Mathematica based interface to the Star-P system. We are working with Interactive Supercomputing to give gridMathematica users the ability to run large scale numerical algorithms in a distributed memory environment," said Roger Germundsson, director of research and development, Wolfram Research. "This new development will offer the HPC community an even greater array of powerful solutions for high-level mathematical challenges."
Star-P OpenConnect Library API
Star-P 2.4 features an enhanced application programming interface (API) called Star-P OpenConnect that lets users easily plug in numerical libraries and solvers to boost productivity. They can plug in existing serial and parallel libraries, work on them with their preferred desktop tool, and execute the code in task- or data-parallel modes. This enables scientists and engineers to tap the vast array of open source and commercial libraries, such as those from Numerical Algorithms Group, which may have previously been beyond their technological reach.
"Star-P enables a whole new class of user to take advantage of our numerical libraries," said Rob Meyer, CEO of Numerical Algorithms Group. "The introduction of cost-effective multi-core processors by AMD and Intel is making high performance computing accessible to scientists, engineers and commercial users. The ability to plug our libraries into Star-P removes the final hurdle to taking full advantage of these computing resources, eliminating the complexities of parallel programming."
The Star-P OpenConnect Library API also enables organizations to breathe new life into internally developed serial libraries that would otherwise languish unused. By automatically parallelizing the libraries, Star-P extends these organizations' intellectual investments in custom library development.
Support for hardware accelerators
Star-P 2.4's open architecture enables users to tap the computational power, I/O and memory of third-party hardware accelerators such as math coprocessors, field programmable gate arrays (FPGAs) and graphics processing units (GPUs). Through the Star-P OpenConnect Library API, compute-intensive algorithms embedded in hardware appear as standard library functions that can be easily accessed within a desktop application.
Star-P solves an important hardware barrier to the widespread adoption of accelerators in HPC, which traditionally have been an unfamiliar environment to scientists and engineers. Star-P provides a higher level of abstraction to users by coupling familiar desktop tools with a high-performance back-end execution engine.
"ClearSpeed's programmable coprocessors enable some of the world's most compute-intensive applications to run on affordable industry-standard platforms including clusters, servers and workstations," said Simon McIntosh- Smith, vice president of applications at ClearSpeed Technology. "Star-P's open support for our acceleration technology will help make high performance computing not only more cost effective, but enable much easier application development."
Pricing and availability
Star-P 2.4 is available immediately with list pricing starting at $7,995 for a 4-socket server license.
About Interactive Supercomputing
Interactive Supercomputing (ISC) launched in 2004 to commercialize Star-P, an interactive parallel computing platform. With automatic parallelization and interactive execution of existing desktop simulation applications, Star-P merges two previously distinct environments - desktop computers and high performance servers - into one. Based in Waltham, Mass., the privately held company markets Star-P for a range of security, intelligence, manufacturing, energy, biomedical, financial, and scientific research applications. Additional information is available at http://www.interactivesupercomputing.com/.
Source: Interactive Supercomputing
CONTACT:
Ilya Mirman
Interactive Supercomputing,
+1-781-419-5088
Web site: http://www.interactivesupercomputing.com/
