The Defense Advanced Research Projects Agency (DARPA) hired three companies to answer the question “Can we really build a useful quantum computer?” Over the next five years, each company will use a unique design theory to see if they can make it work. DARPA announced the partnership and immediately set to work reviewing proposed designs.
The problem with designing a quantum computer is the qubits. While offering an only dreamed of speed for computing, the qubits don’t always behave the way they are supposed to.
“You might have 100 or 1,000 of these noisy qubits working together to collectively create one kind of noise-free computational grade, what we call fault tolerant qubit. And those are the kinds of machines we’re interested in,” said DARPA program manager Joe Altepeter in an interview with Federal News Network.
Each of the three companies, Microsoft, Atom Computing and PsiQuantum Corp., presented a development plan to DARPA with a solution for a utility-scale quantum computer that can be built as designed with practical applications. DARPA described utility scale as having more value when it is completed than the cost of building it.
“They have really exciting approaches, they each have something about their approach that nobody else is doing, that I think is really worth taking a careful look to see is there. So I’m incredibly excited,” said Altepeter.
DARPA wants the vendors to build a fault-tolerant computer, meaning they have to figure out a way to stabilize the qubits.
PsiQuantum plans a computer using photons to stabilize the qubits. It’s an approach that has possibilities, but also some problems. The photons don’t react to electric magnetic fields, but they also don’t react to other photons, according to Altepeter.
“That’s as big of a disadvantage as the first one was an advantage. And so you really need some very creative ways if you’re going to make this work to overcome that seemingly significant obstacle,” Altepeter said.
The Microsoft plan for a quantum computer involves a topological superconducting approach that features individual superconducting qubits. Microsoft said these qubits show more stability than previous types without compromising size or speed, according to a company research blog.
Altepeter said it has promise, but also lots of ways to fail.
The third design from Atom computers uses neutral atoms that get trapped with lasers so they don’t have a charge. It’s a program that DARPA has funded in various forms over the last 10 years without a successful result.
“Maybe this is a really quickly scaling approach that we disregarded too quickly a while ago. And so I think for completely different reasons, we’re really excited about all three approaches we’re looking at” Altepeter said.
Regardless of how the computer is ultimately built, it has to provide value in accomplishing tasks that can’t be accomplished by a traditional computer. In the initial phase of the projects, a DARPA-lead test and validation team of experts from government laboratories and federally funded research and development centers will evaluate the concepts.
“And what we really care about is that we build a quantum computer that is good at something that’s really important that we didn’t have another way to do better. And so I think we are moving into the regime where quantum computers are going to start to do things that are legitimately useful even in this noisy regime. But just because they’re useful doesn’t mean that you couldn’t do it better with a laptop,” said Altepeter.