Look at the results and it’s easy to see why Sandia has received two national awards from the Federal Laboratory Consortium for Technology Transfer for its partnerships with Cray Inc. and Stirling Energy Systems, Inc.
Look at the results and it’s easy to see why Sandia has received two national awards from the Federal Laboratory Consortium for Technology Transfer for its partnerships with Cray Inc. and Stirling Energy Systems, Inc. (Photo by Randy Montoya)
Sandia and Cray joined forces in 2001 to build the Red Storm supercomputer, the predecessor of the Seattle, Wash.-based company’s line of Cray XT supercomputers. In 2009, Jaguar, a Cray XT5 supercomputer housed at Oak Ridge National Laboratory, won the prestigious Gordon Bell Prize for parallel computing performance. And, Franklin, a 350-teraflop Cray XT4 system at Lawrence Berkeley National Laboratory, was ranked 11th fastest in the world, according to the award nomination.
Since starting its collaboration with Sandia, Stirling Energy Systems (SES) has signed contracts to provide 1.85 gigawatts of solar power from its concentrating solar power system, the SunCatcher™. The company also is planning to build the world’s largest solar energy generating system on about 6,500 acres in southern California. The 750-megawatt Imperial Valley Solar plant will power 562,500 homes in the San Diego area by 2014.
The Federal Laboratory Consortium (FLC) presented Sandians with the awards April 29 at its national meeting in Albuquerque.
“Sandia has always done well in these recognition awards and it’s an indication of our ability to transfer technology to industry,” says Hal Morgan, senior manager for Industrial Partnerships and Strategy Dept. 1930.
SunCatcher at Sandia
The SunCatcher™ technology makes it possible for large power facilities to provide clean, reliable, cost-effective, and sustainable solar-electric power to communities, address renewable energy-portfolio targets, and tackle global climate change by reducing carbon emissions, according to a nomination submitted for a Federal Laboratory Consortium (FLC) awards program.
Sandia was recognized recently by the FLC for its tech transfer work with Stirling Energy Systems (SES) and the SunCatcher technology.
Steve Cowman, SES chief executive officer, says Sandia’s relationship with the Scottsdale, Ariz.-based company was essential for attracting investments of about $200 million over the past two years from a firm with which he was previously associated.
“The product has been significantly enhanced and improved by virtue of the collaboration and partnership that we have with Sandia,” Cowman says.
The close-knit collaboration started in 2003, when SES sent a prototype solar dish and two part-time engineers to Sandia’s state-of-the-art National Solar Thermal Test Facility (NSTTF), says Chuck Andraka (6337), Sandia’s lead project engineer. Sandia made space and equipment available, and provided access to technical personnel to help SES develop, test, and refine their dish engine system, he says.
“That was so successful and they learned so much about their own dish from our work that it quickly blossomed into a proposal for a mini-power plant of six dishes,” Chuck says.
In 2006, DOE approved a cooperative research and development agreement (CRADA) with SES.
The openness demonstrated by representatives from both partners allowed technology transfer to flow in two directions, enhancing the skills of Sandia’s researchers while contributing to the development of a commercial product for SES, the partners say.
Sandia’s technical expertise helped SES drop 4,000-6,000 pounds of steel from what was a 16,000-pound structure and halved the number of mirrors from 80 to 40, which reduced construction and maintenance costs.
The Sandians also convinced SES to use a radial design with pie-shaped mirrors, the design the company is producing today, Chuck says.
“It’s much more efficient structurally, it’s quicker to assemble and it’s better optically,” Chuck says. “That was one of the big payoffs of our interaction with them.”
Eventually, 20 full-time SES engineers and technical staff worked at NSTTF. The engineers learned from each other, building and troubleshooting together. Sandia contributed its expertise in solar engineering, particularly in the area of solar optics, but Stirling gave Sandians the opportunity to work directly on real hardware owned by SES, fine-tuning their ideas so they would work in real-world applications.
“It’s a closer collaboration than we’ve ever had before,” Chuck says. “We’re actually sitting in on the design process, working with their engineers and mentoring people.”
In 2008, one of the original SunCatchers set an efficiency record for converting solar energy to electricity by achieving a 31.25 percent net efficiency rate that toppled a 1984 record of 29.4 percent.
The collaboration is ongoing.Gaining a better understanding of optics and speeding up the automatic alignment of the dish’s mirrors so they will work for rapid deployment of large-scale generating plants are areas for further research, Chuck says.
“We’re learning at Sandia whether these techniques are even going to work and we’re developing intellectual property that we’ll be able to market to other companies,” Chuck says.
Cowman and Chuck agree that without the partnership, there would be no SunCatcher.
“If it hadn’t been for Sandia, we would have had a technological curiosity rather than a commercial project,” Cowman says.
Cray and Sandia: Partnering to create Red Storm and successor systems
When Cray and Sandia joined forces in 2001 to build the Red Storm supercomputer, there were no commercial supercomputers that targeted complex simulations, says Sudip Dosanjh (1420), senior manager of Computer & Software Systems.
Sandia was recognized recently by the Federal Laboratory Consortium for its tech transfer work with Cray.
Initially, the customer-supplier relationship between Sandia and Cray was switched. Sandia was unable to find a provider for the supercomputer architecture, so the Labs supplied the architecture and several fundamental technologies and Cray’s engineers jointly created the system with the Labs and built the hardware to Sandia’s specifications, says Ron Brightwell (1423), manager of Scalable System Software.
In spite of that initial role reversal, the two partners built what later became one of the best massively parallel processing systems in history: the Cray XT series, according to the award nomination. These systems have had significant national impact.
“The NNSA’s partnership with Cray has been important to both the nation’s nuclear security and to national supercomputing competitiveness. The government’s investment in Red Storm and the XT line of supercomputers has afforded scientists unprecedented capabilities for exploring long-standing problems,” says Robert Meisner, head of NNSA’s Advanced Simulation and Computing program.
Red Storm’s development took about two and a half years, about a year less than the typical vendor schedule.
Peter Ungaro, Cray’s president and chief executive, credits Sandia for the speed of the development. “We would have gotten there, but we definitely wouldn’t have done it in the timeframe that we got there with Sandia, and we wouldn’t have built as good of a product, if we had done it ourselves,” he says.
The partnership included a Work for Others (WFO) agreement, licenses, and jointly held patents.
Sandia developed several technologies that have been or still are being used in the Cray XT computers, including a Portals networking layer that enables processors to communicate; the Catamount Lightweight Kernel to improve processing and data access performance; and the Compute Processor Allocator that optimizes how processors are assigned to computing jobs.
Red Storm enabled Sandia and the DOE to meet its computing needs by using tens of thousands of processors working in parallel for several weeks on a single problem.
The Cray XT technology has proven effective at solving a wide range of science and engineering problems related to climate change, fusion, material science, nanomaterials, biology, and astrophysics, according to the nomination.
There also have been economic benefits from the partnership. Before collaborating with Sandia, Cray was striving to stay competitive in the volatile supercomputing market, according to the nomination. Since introducing the Cray XT, company statistics show it has sold more than 1,200 XT cabinets to more than 80 customers worldwide.
“The international success of Cray’s XT line demonstrates that there is a strong commercial and government demand for highly scalable systems,” Sudip says.
Hal says the partnership with Cray has provided a much-needed supplier in the high-performance computing market.
Meisner says the DOE partnership with Cray will continue through the purchase of the Cielo platform, which will be jointly managed by Los Alamos National Laboratory and Sandia.
And, the two partners view the Cray-Sandia partnership on Red Storm as a potential model for the next supercomputing challenge: exascale computing, where systems can handle a million trillion operations per second.