INDUSTRY

Since the launch of high-tech startup Interactive Supercomputing (ISC) nearly a year ago, more than two dozen organizations have already adopted the company's Star-P software for solving some of the world's greatest scientific and engineering challenges. Star-P is an interactive parallel computing platform that enables users to code algorithms and models on their desktops using familiar mathematical software packages such as MathWorks' MATLAB and run them instantly and interactively on parallel high performance computers (HPCs). While parallel computers are increasingly vital to scientific discovery due to the growth of computationally intensive problems, recent studies* show that re-programming code to run on these notoriously complex systems is one of the biggest productivity killers of scientific research. Star-P solves this problem by eliminating the need to re-program applications in C, FORTRAN or MPI languages to run on parallel computers. Star-P automatically connects desktop applications to computers and parallelizes the application code on the fly. Consequently, organizations in government, academia, life sciences and other industries are using Star-P to tackle much larger problems on their desktops than ever before possible, while arriving at a solution in a fraction of the time. For example, researchers at Australia's Howard Florey Institute are using Star-P to rapidly analyze large MRI datasets that may someday reveal correlations between brain structure and conditions such as ADHD, Alzheimer's, auditory hallucinations and primal urges (e.g. thirst). The Air Force Research Laboratory (AFRL) in Rome, NY, is developing a new supercomputer-based system with Star-P that could help the military analyze massive amounts of satellite-based radar data for surveillance, missile warning, communications and navigational activities. And the Massachusetts Institute of Technology is using Star-P to predict thermodynamic properties of molecules that impact weather patterns and pollution dynamics. Other Star-P customers include: - Baylor College of Medicine - Cleveland Clinic - Dartmouth Medical School - Duquesne University - General Atomics - M.D. Anderson Cancer Center - Montana State University - National Cancer Institute - National Center for Supercomputing Applications - National Oceanic and Atmospheric Association - University of Connecticut - Pennsylvania State University "Scientists and engineers are looking for easier and better ways to advance their science, which often requires new programming models like Star-P in order to take advantage of HPC cluster systems," said Earl Joseph, IDC Research vice president and executive director HPC User Forum. "HPC cluster systems have become the fastest growing IT sector, with growth rates exceeding 70 percent over each of the last four years, and HPC users are constantly looking for ways to make better use of these systems." Star-P's early adoption by the technical computing community stems largely from it being an open platform that supports easy integration. Software vendors, systems integrators and even customers can plug in a variety of custom and commercial libraries, solvers and hardware accelerators and run them in parallel with Star-P, such as libraries from Visual Numerics and Numerical Algorithms Group. It also supports popular open source libraries like Trilinos from Sandia National Labs. "Signing up dozens of the world's most renowned research labs and organizations in the first year of our operation is enormous validation for Star-P," said Pete Peterson, ISC's president and CEO. "It's bringing a much needed productivity accelerator to researchers who have struggled with desktop performance on increasingly larger and more complex simulations. With Star-P's new programming model, they can stay within their familiar environment, yet tap into the power of parallel clusters."