If you’re a public or private entity looking for “big data” solutions, Lawrence Livermore National Laboratory (LLNL) and IBM are, in a manner of speaking, open for business.
The Department of Energy’s California-based laboratory announced the new arrangement with IBM on Wednesday. The new agreement, called “Deep Computing Solutions,” builds on a roughly 20-year relationship between the two entities. The result: An opportunity for third-parties to leverage the human and digital brain power of LLNL and IBM towards their “big data” needs.
The news comes roughly a week after it was announced that IBM’s Sequoia system, housed at LLNL, was the world’s fastest supercomputer, ranking first on the TOP500 list — the first time since 2009 that a computer in the U.S. took the top spot.
The “Deep Computing Solutions” agreement will bring together IBM Research with LLNL experts at Livermore’s High Performance Computing Innovation Center. As part of the arrangement, LLNL acquired a new system from IBM called Vulcan — a 24-rack Blue Gene/Q based on the POWER architecture. It will be delivered this summer, according to a press release issued Wednesday.
If you’re hoping to use Sequoia, Vulcan’s faster counterpart, forget it. The system is restricted to classified work on behalf of the DOE’s National Nuclear Security Administration (NNSA), whereas Vulcan will be made available for unclassified, third-party work.
And discussions are already underway with “a broad range of American industry collaborators,” according to the Wednesday release. When asked which entities made up that “broad range,” IBM was mum.
“We’re in negotiation with a number of partners right now,” said Jim Sexton, IBM Research Program Director for the Computational Science Center at IBM Watson Research Center. State and federal governments as well as private companies could participate.
“The golden result,” says Sexton, “is an improved process for an agency or a company working with us.”
This means an engineering company, for example, would have a faster design cycle; a company managing large data would have better insight on that data; a plant would have better economies of scale and better returns on investment.
The partnership is currently planned to last five years, but the intention is to make it permanently viable. “If we’re successful,” Sexton said, “we would expect to be copied in many different ways.”
And for those worried that IBM or any third-party entity will be getting a free supercomputer on the government’s dime, not to worry: “There’s no free computing here,” Sexton said. The third party bears the marginal cost of the extra activity.
The announcement will be part of briefing on how government can harness big data to tackle the nation’s challenges, ranging from health care and transportation to fraud, waste and abuse. The briefing will take place Wednesday at 11:30 a.m. on Capitol Hill and feature Sen. Diane Feinstein (D-Calif.) and Rep. Steve Womack (R-Ark.).
Read more news and ideas on Innovations: