What will quantum computing actually look like?

The field of quantum computing, like qubits themselves, exists in a state of indeterminacy, with multiple companies making big claims about wildly different approaches. On Thursday, DARPA—the Defense Advanced Projects Research Agency—launched its latest effort to sort practical approaches from chimeras, handing contracts to 15 companies that will attempt to benchmark their work. 

So little is known about the field that there is no way—at least, not yet—to predict what a truly “useful” quantum computer would look like, how it would perform, what materials or power it would need, or how much it would cost, said Joe Altepeter, program manager for the DARPA Microsystems Technology Office. 

“There’s still room for none [of the different approaches] to work, or seven totally different technologies to work. It’s like comparing vacuum tubes to transistors to abacuses. There’s completely different models for computation, and in a year or two, we’re going to know a lot about which ones can work hard and which ones can’t,” which is why the program exists, he said.

Altepeter said most of the customers who are buying time on today’s quantum computers are simply doing research, not seeking practical applications. The agency wants a way to measure the performance of one approach over another, especially when applied to an actual real-world challenge. 

“We don’t know what the metric is, and as far as I can tell, all of the quantum computers that exist today don’t do anything industrially useful. Many of them are scientific tours,” he said. 

What would be industrially-useful? Any problem that requires far more calculations than can easily be solved by modern, binary computers, such as highly-complex chemical reactions, which could lead to new types of rocket-fuel and ship coatings to prevent corrosion. Figuring out those applications involves actually talking to companies about the difficult problems they face in discovering new things in areas such as rocket fuel, anti-corrosion and magnetic memory. Quantum computers also promise to break some of the key encryption protocols that support secure communications, whether between a military unit and headquarters or a customer and his bank account at an ATM. (None of the recipients are doing cryptanalysis, but that threat highlights the importance of better understanding quantum computing, he said.)

Altepeter emphasized that the recipients aren’t in competition with one another. Each brings a unique approach or capability that will shed light on what quantum computing will actually look like.

During the program’s first six months, he said, each company will get $1 million to detail their concept and how it might lead to a computer that can do substantial work. Some recipients will then get $15 million to test their ideas with government labs.

Some 300 people from 10 labs will be involved, Altepeter said.

“We’ve got Oak Ridge National Labs, Sandia National Lab, Los Alamos National Lab” and numerous others, he said. “Nothing like this has ever been built before.”

The third phase will be to design the prototype and validate it—with potential awards of $300 million. 

If it’s successful, the program will answer quotations about the future shape of a quantum computing industry, where there may be supply-chain challenges, what the workforce might consist of, what the cost of systems could be, what it might useful for near-term and far,

The recipient list includes well-known names in quantum computing such as IBM as well smaller companies like Photonic and Alice and Bob across the United States, Canada, Australia. An additional three recipients may be named following negotiations.