DARPA Wants to Explore Practical Tech Impact of Quantum Computing

Quantum computing is not a technology widely used in the federal government today, but it could be in the future. The Defense Advanced Research Projects Agency lives in the future, and it wants to know more about the technology now so that it can help the Pentagon get ahead of China and other adversaries.

The Defense Department’s research arm last month issued a request for information on quantum computing, seeking input on how the technology will practically impact other areas, including artificial intelligence, physical systems and data analytics. Specifically, DARPA’s Defense Sciences Office is looking for information “on new capabilities that could be enabled by current and next generation quantum computers for understanding complex physical systems, improving” artificial intelligence, machine learning and enhancing distributed sensing. Notably, DARPA wants to explore quantum computing’s effects on “hard” science and technology problems and not cryptography.

DARPA’s outreach to industry and academia comes as Congress contemplates ramping up funding for quantum research. DARPA says in its RFI that it wants to challenge the scientific, academic and private sector communities “to address the fundamental limits of quantum computing and to identify where quantum computing can relevantly address hard science and technology problems.”

The RFI seeks responses on four key challenges: the fundamental limits of quantum computing, hybrid approaches to machine learning, interfacing quantum sensors with quantum computing resources, and quantum computing-inspired algorithms and processes that are applicable to classical computers.

Responses are due Aug. 10 and will be used to identify participants and speakers for a potential workshop.

DARPA Wants to Understand How Quantum Computing Will Impact AI 

First, some basics. What is quantum computing? As FedTech has reported:

Quantum computing harnesses the laws of quantum mechanics to carry out complex data operations. While traditional computers use bits (represented as either binary 1s or 0s), quantum computing harnesses quantum bits, known as qubits. These can be read as 1s, 0s, or both, providing exponential computing power over traditional computers by creating shortcuts in the computing process. The challenge, though, comes in scale. The more qubits a machine uses, the more likely a breakdown will occur.

DARPA is concerned about how quantum-computing capabilities will develop and be limited over the near term (the next few years) and longer term (the next few decades).

An issue of particular interest to DARPA is the potential impact of quantum computing on so-called “second wave” AI/ML optimization. Machine learning, DARPA says in the RFI, has “shown significant value in a broad range of real world problems, but the training time (due to the size and variety of the data needed for learning) and also network design space (due to a paucity of detailed analysis and theory for ML/deep learning (DL) systems) are large.”

Quantum computing has the potential to “significantly decrease training time of currently standard ML approaches by providing quantum speedup on optimization subroutines,” DARPA notes in the RFI. New ML capabilities could potentially be unleashed by combining a limited number of quantum computers with either existing quantum sensors or classical computing resources, which might “bypass the problems of state preparation and interfacing to classical memory.”

“In this case it has been posited that by aggregating quantum data from distributed sensors, a quantum computer may improve the performance beyond what could be classically achievable,” DARPA says.

DARPA also seeks information on “adapting to classical computers some of the techniques that are being developed for handling quantum data (both at the algorithm level as well as protocols for loading, storing and transferring data).” Such “quantum inspired” approaches may “provide novel capabilities in terms of efficiency and speed,” DARPA notes.

Congress Aims to Ramp Up Quantum Research Funding

The U.S. is not the only country interested in and investing in quantum computing. Indeed, China seeksto be a leader in the field, and “its government is building a $10 billion National Laboratory for Quantum Information Sciences in Hefei, Anhui Province, which is slated to open in 2020,” Bloomberg News reports.

Last month, Sen. John Thune, chairman of the U.S. Senate Committee on Commerce, Science and Transportation, said in a statement that the U.S. “is now in a race with China and Europe to develop the next technological breakthroughs based on the power of quantum science. It's a race we must win.”

Thune and Rep. Lamar Smith have introduced companion measures in the Senate and House, the National Quantum Initiative Act of 2018, to “help align and accelerate public and private research and development of quantum science,” as Thune’s statement puts it.

Back in late June, the House Science, Space, and Technology Committee unanimously approved its version of the bill. The bill would establish a National Quantum Coordination Office within the White House Office of Science and Technology Policy “to oversee interagency coordination, provide strategic planning support, serve as a central point of contact for stakeholders, conduct outreach and promote commercialization of federal research by the private sector,” a statement from the committee says.

Importantly, as Science magazine notes, the bill would authorize three agencies — the Energy Department, the National Institute of Standards and Technology and the National Science Foundation — to together spend $1.275 billion from 2019 to 2023 on quantum research. DOE would get $625 million of that total, NIST would receive $400 million and NSF would get $250 million.

This content is made possible by FedTech. The editorial staff of Nextgov was not involved in its preparation.