AEROSPACE

Workshop Reveals Market Forces and Technical Challenges Hindering Development of New and Innovative Application Software; Failure to Improve Situation May Inhibit U.S. Competitiveness -- The Council on Competitiveness today announced the results of the first of its kind industry report on the state of application software for high performance computing (HPC) systems. The findings, taken from a workshop conducted in 2005, reveal that a variety of market forces and technical challenges in recent years have caused independent software vendors (ISVs) to turn their focus away from creating new and innovative HPC application software. As a result, U.S. companies often do not have the production quality HPC application software they want and need. The Council warns that unless this situation is reversed, U.S. companies could be at risk if other countries or companies capitalize on the potential of HPC first. "Advanced, easy to use HPC application software is a 'must have' if we are going to harness the full potential of our high performance computing assets for economic gain," said Council on Competitiveness President Deborah L. Wince-Smith. "While the Council observed many barriers impeding the development of this critical software, we also saw opportunities to leverage public and private sector investment and expertise in HPC to address this challenge." On July 14, 2005, the Council of Competitiveness and the Ohio Supercomputer Center, with support for the Department of Energy's Office of Science, co-sponsored a daylong workshop to examine the reasons behind the current lack of production quality HPC application software. Workshop participants included HPC users, ISVs, university researchers, hardware vendors, and government scientists and engineers. They noted that the HPC market remains a niche within a much larger commercial computing market, inhibiting the development of the HPC application software the industry needs to achieve competitive advantage. At the end of the workshop the participants agreed on a number of recommendations for ISVs, universities and the federal government that could stimulate growth in the HPC market and overcome barriers preventing development of advanced HPC application software, including: * HPC ISVs, in consultation with their customers, should consider alternative business models in order to increase revenue and financial stability, and encourage HPC market growth. * The nation's universities should increase their educational programs in computational science at the undergraduate and graduate levels to meet the need for skilled technical workers. * The university and laboratory research communities should enhance their understanding of ISV needs and requirements so they can leverage their own software research and education agendas to assist ISVs where appropriate. * The government should modify its research support practices to provide sustained (multi-year) funding for research teams to develop mature research software and algorithms, and should encourage ommercialization of suitable software. A complete list of recommendations is included in the full workshop report available at its Web site. High Performance Computing Initiative The Council's High Performance Computing Initiative has galvanized a dialogue among government agencies, system and software developers and private sector users of high performance computing to leverage government R&D investment in this technology and facilitate wider usage across the private sector to propel innovation and competitiveness. For more information on the Council's High Performance Computing Initiative and its annual HPC Users Conference please visit its Web site. What the market is saying: "We're hoping to work at two tiers: at the bottom, to increase awareness and simplify usability for [everyday] people... and then again at the top, where you have relationships with national labs and universities." - Wood Lotz, CEO, Absoft "Over the years, we've viewed hardware as an enabler, and the breakthroughs come from application software that can leverage the latest hardware." - Stan Posey, HPC Industry and Business Development, Silicon Graphics, Inc. "We can't say 'niche' often enough. I think this community sometimes misunderstands what a niche is...Our challenge is further exacerbated by the fact that this is a very small market, with a multiplicity of requirements that are disproportionate to its size." - Chris Doehlert, President & CEO, Etnus LLC "It's a matter of how you cast competition...small, private companies [worth] five or ten million dollars are eminently acquirable without constraint by anybody in the world. I can't stop that, so I have to exist with it." -David Turek, Vice President, Deep Computing, IBM "The analogy is you're going to rent a car for the weekend. You go to Hertz...and they say, sorry, all we offer are annual leases. But you only need the car for the weekend. Sorry, they say again, all we offer are annual leases. That is the ISV community today. It needs to change. We need to be asking, 'How much do you need and how long do you need it?' " - Paul Bemis, Vice President Marketing, Fluent Inc. " We're not used to setting aside millions of dollars a year to build HPC systems. We would rather go hire the next Ph.D. to extend the reach of Virtual Product Development and address the scalability requirements of our solutions." - Dr. Reza Sadeghi, Vice President, Enterprise Computing, MSC.Software Corporation "I think the fundamental problem is that we're all driven to using the cheapest system, which today is some variant of a PC-cluster ...ISVs have to support that. And that makes it almost impossible for them also to design for a better system." - Dr. Robert Lucas, Computational Sciences Division Director, University of Southern California Information Sciences Institute The people we need to architect these codes are unique individuals - they have to understand the science domain, computational science, computer science, and then figure out how to make these big applications run on a parallel computer and to scale." - John Morrison, Division Leader for Computing, Communications & Networking, Los Alamos National Laboratory "Start digging as to where these software packages came from, an awful lot of them have their roots in government-funded development that was commercialized. [But] one of the things in my view that is institutionally different now than it was 30 years ago or even 20 years ago - the amount of resources going into those kinds of activities has been steadily drying up." - Dr. Phillip Colella, Senior Staff Scientist, Lawrence Berkeley National Laboratory "I believe there are literally thousands of new applications out there that do things that the current ISV applications aren't yet capable of. Yet, it appears to be extremely difficult on both sides to move those applications into the ISV community." - Dr. William Camp, Director, Computation, Computers, Information and Mathematics, Sandia National Laboratories "Huge data-intense problems are going to change the balance of hardware and software architectures." - Dr. William Camp, Director, Computation, Computers, Information and Mathematics, Sandia National Laboratories "When I look at the successful packages that have come out [of the national labs], they all came out with a BSD type of open source license." - Al Geist, Corporate Fellow, Oak Ridge National Laboratory "Legacy code does put a huge burden on us for backward compatibility." - Wood Lotz, CEO, Absoft "How do you get applications into the hands of people who are not going to be the experts? That raises the general question of, how do you set the balance between functionality and interfaces? And when does one trump the other? I think a lot of people might say that the application with the best interface will win almost every time because of the fit into the workflow." - Dr. Donald Paul, Vice President and Chief Technology Officer, Chevron Corporation "If I were an ISV, the first thing I would do if I were going to bring out a first product is make sure that I had my application interface characteristics welldefined so it could plug into any user interface that a client may have already installed locally," - Dr. George Michaels, Associate Laboratory Director, Computational and Information Sciences Directorate, Pacific Northwest National Laboratory "We shouldn't underestimate the value to the ISV if you can make a service model work, because what the ISV is doing over time is accreting a disproportionate share of the intellectual capital, two-legged capital. Pretty soon, the ISV has a pool of talent that's connected to the technology... they know how it's used. They know what problems count. They know what the solutions are worth. That collection of knowledge reaches some critical point and you really have something else you can sell." - Dr. Donald Paul, Vice President and Chief Technology Officer, Chevron Corporation "The availability of easy-to-use, shrink-wrapped HPC software will lead to profound gains in commerce, industrial productivity, scientific discovery, and greater national security." - Dr. Stanley Ahalt, Executive Director, Ohio Supercomputer Center "By leveraging the investment and expertise of both the public and private sectors, we can harness the innovative capacity of the high-performance computing community and lift our country to a new level of competitive advantage." - Dr. Stanley Ahalt, Executive Director, Ohio Supercomputer Center "In each of these cases [where software made the transition from research to production quality HPC applications] there was a long term investment by the government to give these packages a chanced to mature." - Al Geist, Corporate Fellow, Oak Ridge National Laboratory "The government budgets aren't big enough to do it alone. The industry is under ever more competitive global pressures. The start-up companies can't find the venture capital to invest in this space and so partnerships, to me, seem to be the only viable way out." - Scott Metcalf, President and CEO, PathScale, Inc. "I don't think that it's always the case that blindly following open source is necessarily in the country's best interest in terms of building a competitive structure. And I say that as a member of the Free Software Foundation, who makes a lot of money off of both BSD and GPL code. I'm an insider, but I'm still going to throw rocks at it because it's not always in our national interest." - Scott Metcalf, President and CEO, PathScale, Inc.