The Leading Source for Global News and Information Covering the Ecosystem of High Productivity Computing
June 15, 2007
The keynote topic at TeraGrid '07 was cyberinfrastructure, as you'd expect. But if you walked into the University of Wisconsin's Union Theater on the morning of June 5 expecting technotalk about bits, bytes and FLOPS, you were in for a surprise.
The speaker, Anita Jones, spent the past several decades helping to guide the direction of fundamental research in the United States. She is accustomed to thinking from the perspective of policy, and her talk -- "Stable High End cyberinfrastructure," subtitled "in the presence of revolutionary changes in software, hardware and usage" -- offered a bird's-eye view of cyberinfrastructure since the 1970s, what has changed and what hasn't.
Few people can speak about research policy with the authority Jones brings. Director of Defense Research and Engineering for the U.S. Department of Defense from 1993-97, she has also served on research-related boards and panels for the U.S. Air Force, NASA, the National Research Council, the Department of Energy and the National Science Foundation. Especially to the point for the TG '07 audience, she was vice chair of the National Science Board in 2001 when it first approved the TeraGrid.
She began with a brush-stroke summary of how U.S. civilian research computing has evolved from the 1970s until today. The 70s heralded parallelism. Fifty processors were the most around, but that was enough, Jones said, to learn that parallelism was a mind-bending concept. "It takes a long time to get your head around concurrency."
The 80s brought a major change in the science and engineering landscape with creation of the NSF supercomputing centers; for the first time, grant researchers had access to HPC. The 90s saw access to HPC extended via the transition to NSF's PACI (Partnerships for Advanced Computational Infrastructure) program. These two policy emphases -- increasing "high end" resources and "broadening access" to users -- compete with each other for resources, said Jones, and today, through the TeraGrid and the NSF cyberinfrastructure program, there's a renewed awareness about extending access more broadly.
The measures of the computational research investment have also evolved, Jones noted, from the number of machines to peak cycles to computational performance on benchmarks. The next set of measures, however, yet to be well defined, should include more than cycles delivered to the computation and should take into account other factors, such as how effectively memory is used and the usefulness of visualization.
In allocating the scarce resources of taxpayer dollars, said Jones, it's important to remember that "you can't shoehorn knowledge and experience into people." People expertise is what matters most. After that is software. These two are the first priorities to consider when managing how you invest in hardware. "When lead scientists have funds to hire the next post-doc," she said, "in almost every case they hire a discipline scientist. This is often a mistake, because you have less than first-class knowledge about how to do the computation."
The emphasis on people is also key with respect to "the time machine metaphor," a notion first described in the Hays Report in 1994 and which Jones stressed at several points in her talk. "You want to give some people the tools to catapult into the future compared to the rest of us. They will develop new algorithms, new discontinuous approaches to solving problems. This is the high end."
Jones distinguished the "high end" from the "middle end" -- more broadly accessible resources. "For the health of the research community, there needs to be a select set of people. The nation can't afford to let everyone live in the future, and this select few enable the assets that will define and pay for the future." The time machine metaphor, she said, holds not only for "big iron" but also for visualization and database management.
NSF, Jones observed, has to work out the question of how to balance between broad access and the high end. "This will continue to require centers," she emphasized, "because it's important to keep high-end computing at educational sites. People, remember, are number one."
Jones served on the Committee on Science, Engineering and Public Policy that produced Rising Above The Gathering Storm: Energizing and Employing America for a Brighter Economic Future (2007), the recent report that ties U.S. innovation and economic health to, among other things, U.S. success in high-end computing. This report led President Bush to put his "competitive initiative" into his State of the Union address two months after the report came out.
Jones stressed that in competition one looks for an "unfair advantage" and high-end computing offers one. "If there's anywhere we can innovate," said Jones, "it's at the high end. U.S. taxpayers pay for 99-percent of the research in this room," said Jones. "You want the U.S. economy to prosper because it leads to more support for basic research. And that in turn lays the groundwork for the next round of innovation."
In cyberinfrastructure, decisions involve interdependence, and TeraGrid exemplifies this. "TeraGrid knits together scarce assets." Decisions on standards, protocols, data representation and the design of community tools are crucial, and because decisions need to be made, said Jones, it's important to have leadership and not governance completely by consensus. A task for TeraGrid, she said, is to define "Return on Access." "It's not just cycles and fast networks. We live in a richer world. It includes access to storage repositories and visualization."
In closing, Jones encouraged her audience to "keep on pushing the envelope."
(Digg, Technorati, more)
Platform HPC Workgroup Manager
Platform HPC Workgroup Manager integrates all the cluster productivity tools you need to deploy, run and manage your HPC environment.
The ACM Turing Award goes to the creator of the modern personal computer; and Voltaire announces a mid-range InfiniBand switch and new technology that accelerates distributed applications. We recap those stories and more in our weekly wrapup.
Read More...
The prospects for virtual SMP technology got another boost last month when Florida State University announced it had installed a new HPC system from 3Leaf Systems. The servers are being housed at the university's HPC facility and will be used across a range of scientific disciplines.
Read More...
For the first time in 62 years, the four-man Olympics bobsled team from the US captured the gold medal, setting a course world record in the process. The winning bobsled had some state-of-the-art engineering behind it, including CFD software from Exa Corporation. As it turned out, that software may have proved to be the margin of difference in the race.
Read More...
Mar 11 | Linux Magazine | CUDA may be the rage, but OpenCL is a standard that has some features you may need. Read more...
Mar 09 | Free Software Magazine | Data-driven computing will need open software. Read more...
Mar 09 | Bio-IT World | Tahoe Informatics founder eyes GPUs, CUDA software. Read more...
Mar 08 | Sporting Life | Formula One engineers differ on benefits of CFD. Read more...
Mar 08 | InfoWorld | AMD offers up 48-core server prize. Read more...
Jan 12 | | In-depth look at vSMP Foundation server virtualization technology, technical implementation, use cases and capabilities. The technical whitepaper provides an architectural overview and details on the three vSMP Foundation products: vSMP Foundation for SMP, vSMP Foundation for Cluster and vSMP Foundation for Cloud.
Jan 18 | | This white paper discusses Gore’s copper cable assemblies, and how they continue to exceed the standards for providing reliable, cost-effective solutions for high-performance computer applications.
Join this online panel discussion for live Q&A with leading industry experts, analysts, and end-users to discuss the latest innovations, best practices, barriers to implementation, and measurable benefits of server virtualization with a particular focus on today's real world solutions.
Learn about scalable fault-tolerant architectures and examples of energy efficient and scalable supercomputing clusters using dual QDR InfiniBand to combine capacity computing with network failover capabilities with the help of programming languages such as MPI and a robust Linux cluster management package.
LIVE@SCO9: The IBM team discusses new innovations in hardware, software and services that help clients better understand their workloads and get insight from their R&D efforts. Technology demonstrations include the soon-to-be-released Power7 HPC processor, the DCS990 system with 2.4 petabytes of storage, the xCAT management tool, secure HPC cloud computing and more. Winners of two HPCwire Readers' and Editors’ Choice Awards! Take the IBM virtual tour at SC09 or more information go online to: http://www-03.ibm.com/systems/deepcomputing/sc09.html