Thanks to the NSA, quantum computing may some day be in the cloud
- 04 January, 2014 15:27
The National Security Agency (NSA) is spending some $80 million in basic research on quantum computing. And what the NSA spends its research money on may ultimately help commercialize quantum computing -- and even make it accessible via the cloud.
This is what Defense Department agencies do: They fund basic research that private industry sees as too risky, but if the work leads to breakthroughs, it's the commercial sector that may benefit the most.
Information about the NSA's quantum computing research comes from documents acquired by Edward Snowden and published by the Washington Post. These documents use macho sounding names for the quantum computing effort, including "Owning the Net (OTN)."
Although the revelation may create another uncomfortable Big Brother moment for the NSA, the more important question is whether it -- and the government as a whole -- is spending enough on quantum computing.
The U.S. isn't alone in this area. The U.K. government recently announced plans to spend $444 million to create five quantum computing centers. Canada's Institute for Quantum Computing is more than a decade old. China, Russia and the Europeans are also all investing in quantum research.
"It's an academic race at the moment," said Earl Joseph, an analyst at IDC who noted that defense agencies have been funding quantum computing efforts for over a decade. "The goal is to fund basic research and make new discoveries that may be useful for our safety and national defense."
Christopher Willard, chief research officer at Intersect360 Research, sees the NSA spending as a partial recognition that Moore's Law is winding down and is no longer delivering regular improvements in processor cycle time.
If the ability to assemble high performance systems from commercial, off-the-shelf technologies is waning, "the market should then be entering a new phase of experimentation, and computer architecture innovations," said Willard. Quantum computing is but one example of this shift.
It's hard to estimate just how much the U.S. now spends on quantum-related research because the funding comes from multiple agencies and is otherwise classified. But in May, Google, NASA and the Universities Space Research Association began collaborating on quantum computing research and are using a quantum computer developed by D-Wave Systems, the only manufacturer of such computers today.
The type of system that D-Wave has developed has broad applicability to big data problems and analytics, and in areas such as machine learning, said Vern Brownell, the CEO of D-Wave. For instance, Google is interested in using artificial intelligence techniques to develop better algorithms for image searching. It can be used for financial modeling, cancer research and solving difficult logistical problems, he said.
Brownell said that quantum computing will ultimately work as a type of co-processor, and not as a direct replacement for classical computing systems.
In "our view of the future," said Brownell, "there will be quantum computing resources in the cloud." It will be a level playing field, available to any developer who wants to solve a particularly difficult problem, such as machine learning.
Joseph, as well, sees quantum computing operating in tandem with classical computing -- with potential as a cloud-based resource. Quantum computing can quickly provide ranges and probabilities, but to narrow the data down will still require a classical system.
An ordinary computer uses bits that are either 0 or 1. A quantum computer uses subatomic particles that can hold states, 0 and 1 simultaneously. Instead of doing a calculation one after the other, the processing can increase exponentially. Two qubits can hold four distinct states, and 10 qubits can hold 1,024 states.
There are different approaches to building quantum computers, but no agreed upon methods. A lot of research involves developing a system that can utilize Shor's algorithm, named after MIT researcher Peter Shor. That algorithm can help with code-cracking.
Quantum computing isn't a replacement for high performance computing or exascale, despite its potential speed, said Steve Conway, an analyst at IDC. He believes the biggest industrial use for quantum computing will be to improve computer security.
"You will never see a quantum computer doing word processing," said Conway.
Patrick Thibodeau covers cloud computing and enterprise applications, outsourcing, government IT policies, data centers and IT workforce issues for Computerworld. Follow Patrick on Twitter at @DCgov or subscribe to Patrick's RSS feed. His e-mail address is email@example.com.
Read more about high performance computing in Computerworld's High Performance Computing Topic Center.