Engineering Tomorrow’s Quantum Computers

Researchers have touted the revolutionary potential of quantum computers to take on otherwise intractable challenges, like modeling complex molecular behavior for drug discovery or factoring enormous numbers in use for cryptography schemes. But what would a large-scale quantum computer actually look like?

That turns out to be a difficult question to answer, even in general terms. Add parameters and specifications and it becomes even more daunting.

But with help from the Johns Hopkins Applied Physics Laboratory (APL) in Laurel, Maryland, the Defense Advanced Research Projects Agency (DARPA) is attempting to do just that: define, in practical terms, what a useful quantum computer looks like and how it must be built.

Announced in 2022, DARPA’s Underexplored Systems for Utility-Scale Quantum Computing (US2QC) program aims to evaluate less-taken roads to determine whether quantum computing at scale might be achieved in a matter of years rather than decades, as many experts predict — and if so, how. In July 2024, DARPA announced that US2QC will be combined with a related program to found the Quantum Benchmarking Initiative (QBI), which builds and expands upon US2QC.

Joan Hoffmann, who leads the Research and Exploratory Development Mission Area at APL, expressed excitement at the Laboratory’s role in this effort.

“The earliest computers were massive mechanical calculators, and then from there the field went to using vacuum tubes, but what ultimately pushed computers forward was the discovery of transistors,” Hoffmann said. “No one knows what the equivalent of the transistor will be for quantum computing, but it’s fascinating to dig in and try to figure that out.”

More Qubits, Fewer Errors

The beating heart of quantum computing is the quantum bit, or qubit. Unlike classical computer bits, which are coded as either zeros or ones, qubits can exist in multiple states at once. The combination of a qubit’s analog nature — its ability to take values other than exactly zero or one — and entanglement, which connects multiple qubits in nonintuitive ways, allows for novel and extremely efficient solutions for certain challenging problems.

But these very properties also make qubits extremely sensitive to their environment, and thus error-prone.

The quest for a utility-scale quantum computer can be summarized in four deceptively simple words: more qubits, fewer errors. Accuracy demands two more: millions more qubits, far fewer errors.

Therein lies the rub.

“The two current leading technologies — superconducting qubits and ion traps — have shown incredible performance at the single qubit level, but it’s not known whether their dominance will continue as larger fault-tolerant algorithm-scale designs become the focus of the field,” said Scott Hendrickson, APL’s US2QC project manager. “There are many hard problems that have to be solved in order to scale these technologies toward millions of qubits.”

The solution, ironically, may come from technologies that academics threw out long ago, such as photonics, topological quantum computing and trapped neutral atoms: approaches that were slow to mature for making systems with small numbers of qubits, Hendrickson says, but may turn out to be much better for creating systems with millions of them. And commercial companies are now picking up on that advantage.

In January 2023, DARPA selected three companies and tasked them with presenting a design concept for a utility-scale quantum computer. In February, after almost two years of rigorous technical analysis of the design concepts, two of those companies were selected to advance to the program’s validation and co-design phase; and in April, the agency announced that nearly 20 quantum computing companies have been chosen to participate in the initial stage of QBI, in which they will characterize their unique concepts for creating a useful, fault-tolerant quantum computer within a decade.

Engineering Not-Yet-Existent Systems

DARPA tapped APL at the very beginning of the US2QC program to leverage its world-renowned systems engineering expertise and distinguished quantum information team — qualities that uniquely position the Laboratory to create frameworks for evaluating quantum computing platforms using these previously discarded modalities.

But this effort places a paradoxical demand on that expertise, applying systems engineering principles to systems that have not yet been invented.

“If you’re building a skyscraper, it’s all about keeping everything on track, but if you’re going to the Moon for the first time, you don’t know enough at the outset to know what ‘on track’ even looks like,” said Scott Simpkins, APL’s US2QC systems engineering lead. “The moon shot is APL’s strength, and in this context, systems engineering takes on a dynamic, proactive quality.”

For an effort that involves so many unknowns, Simpkins said, that means asking the right questions, which in turn means assembling the right team — in this case, systems engineers and quantum physics experts from across APL.

By combining the results-oriented skills of the former and the relentlessly inquisitive nature of the latter, the team is developing parameters, specifications and performance targets against which to measure the prototype systems being developed by commercial companies.

“Physicists like to attack the most exotic and interesting parts of the problem, but sometimes the reason the technology won’t advance is something very mundane, like not being able to engineer a certain necessary part,” Hendrickson said. “By putting the scientists and the systems engineering into conversation, you get that complete picture.”

Ultimately, if the companies succeed in developing their prototypes and demonstrating their scalability according to the APL-produced standards, the Laboratory team will produce documents that explain — precisely and in depth — how to build, operate and maintain utility-scale quantum computers for specific applications.

“Maybe the government needs one system to be deployable, another that can process sensitive data and still another that needs to be optimized for raw computational power,” Simpkins said. “If all goes well, our work will allow them to create purpose-built systems that meet their needs, whatever they might be.”

Pushing the Field Forward

Colin Trout, a quantum information theorist and lead US2QC scientist at APL, said even from a purely scientific perspective, the application of systems engineering principles may be just what the field needs to move forward.

“Quantum computing systems are getting larger and larger, and building a viable quantum computer is starting to look more like a systems engineering problem than it did even a few years ago,” Trout said. “The lessons we’re learning will be of tremendous value to the field as we work out how to mature these technologies and where they can make the most impact.”

Echoing Trout’s comments, Hendrickson said that the playbook that APL is developing for US2QC is extendable well beyond the current project, even beyond the nascent Quantum Benchmarking Initiative.

“We’re creating an efficient methodology that, if all goes well, we could repeat and apply across a broad range of emerging science and technology areas,” he said.

DISTRIBUTION STATEMENT A. APPROVED FOR PUBLIC RELEASE: DISTRIBUTION IS UNLIMITED.