As those of you who listened to the interviews with the winners of the Itanium Solutions Alliance 2007 Innovation awards (in our May 30, 2007 post), innovation at the highest pinnacles of “high tech” computational development continues to roll on.
In addition to those who won the overall awards in this exciting contest, there were also of course many outstanding finalists with stories that may also be of interest to regular listeners to the Stranova podcast series. Today we are pleased to announce we’ve just posted three new ones:
“UC Riverside’s Department of Computer Science and Engineering”, with Dr. Walid Najjar.UC Riverside was named a Finalist in this year’s Itanium Solutions Innovation awards, thanks to development of a creative means of rapidly accelerating certain types of scientific code used in bioinformatics modeling. These included what is known as the Smith-Waterman algorithm, an intense code set used to measure matching between DNA or protein strings, and the molecular dynamics code NAMD, which helps model how molecules “fold” based on a detailed micro-analysis of atomic forces. To accelerate this modeling, they used their own unique code and what’s known as the ROCCC compiler, in conjunction with a unique SGI server architecture incorporating direct access to Field-Programmable Gate Arrays. The end result was a performance speed increase of over 1000 times what previous “desktop-limited” approaches had previously allowed.
“Interactive Supercomputing”, with Bill Blake, Chief Executive Officer. This company was named a Finalist in the Humanitarian Impact Innovation category of the Itanium Solutions Alliance awards for its development of the Star-P software system and hardware supercomputer installation, all in support of the National Cancer Institute’s latest research in genomics analysis. What Star-P makes possible, through a combination of a desktop client and a server client working together to link supercomputer power to desktop applications, is the ability to run programs like MatLab and Python (which typically run only on high-end desktop systems) on high-performance supercomputer systems without having to do any significant applications reprogramming. The National Cancer Institute’s work using this technology has achieved over 200 times speedup of its applications compared to previous approaches.
“Kindred Healthcare”, with Daniel Poff, Director of Database Administration and SAP Basis at Kindred Healthcare. Kindred Healthcare is a major Fortune 500 healthcare services company, boasting annual revenues of $4.5 billion and providing services in approximately 600 locations in 38 states. Between it and its subsidiaries, it provides that healthcare through operation of long-term acute care hospitals, nursing centers, institutional pharmacies, and the Peoplefirst Rehabiliation Services positioned all around the United States. It has also grown rapidly, with increased complexity of business operations, reorganizations, acquisitions, and the growth of internal databases from 200 GB to 2 Terabytes in ten years’ time. The story Mr. Poff tells is how his team managed the incredible transition in moving his enterprise applications from an older server architecture to one using the latest of hardware and software technology, and doing it all in record time.
You can listen in to each of these by clicking on the following links:computing education human genome itanium Itanium Solutions Alliance supercomputer