The specter of a quantum computer powerful enough to break today’s toughest codes has haunted militaries for years. An Australian project could bring their fears to life by 2027.
Earlier this week, the Australian government inked a $620 million contract with PsiQuantum to create what the Silicon Valley startup says would be the “world’s first utility-scale, fault-tolerant quantum computer”—a building-sized device in Brisbane that just might be powerful enough to break the modern public-key encryption schemes that protect personal data and national secrets.
It’s a rare example of a government placing a “singular bet on a single private quantum computing company,” while also moving away from strictly research-based efforts, said Pete Shadbolt, PsiQuantum’s chief scientific officer and co-founder.
Quantum computers exist already; IBM just installed one on a college campus. But with the help of the Australian government, PsiQuantum aims to be the first to build one that can deliver on the technology’s purported promise of solving math-intensive problems, like finding new drug combinations, that would take a regular computer hundreds of years.
But the only way to see if quantum computing can make good on those promises is to build one without the research-and-development safety net, Shadbolt said.
“Lots of people have small quantum computers,” he said. “They have a small quantum computer that’s scientifically interesting. But the general consensus of the field is that if we’re going to reap the science-fiction-like rewards of a quantum computer, we need a really big one.”
“And for the majority, not all, but for the majority of approaches, that looks like a data center or high-performance computer. It’s a building-scale, large facility, as opposed to some benchtop experiment.”
PsiQuantum plans to break ground on the new facility next year with the ultimate goal of having a working error-correction quantum computer by 2027. But first they must do more work on the quantum architecture, and develop more efficient algorithms and applications, Shadbolt said.
Lawrence Gasman, the founder and president of Inside Quantum Technology, an analyst firm, said the announcement could be a “big thing, because you hear all these wonderful things about what quantum computing can do” and less about practical results.
One of the biggest challenges is having enough qubits, the building blocks of quantum computing.
“Qubits are sad, pathetic little things, and they don’t last long in any physical manifestation. So you need a lot of control. What you want is error-controlled qubits so that you can use a lot more of the qubits, you do have to actually do calculations. And this is arguably the hottest topic in quantum at the moment,” Gasman said. “There’s probably half a dozen companies that have made announcements on error correction.”
Gasman said PsiQuantum’s technical approach uses optical components, which can theoretically use off-the-shelf parts originally designed for telecommunications or data communications.
“I think that out of the four or five viable ways of doing quantum computing, optical computing is a little bit behind. Potentially, it could bounce out to be a very good thing,” Gasman said.
“We’ve come faster to real applications, I should say, than I imagined. And it’ll come even faster if we can get good error collection, and it wouldn’t surprise me if we do, there’s lots of different ways of doing it.”
The U.S. government has been warning of the need to prepare for quantum computing with the Defense Department funding efforts, including the Defense Advanced Research Project Agency’s , or DARPA, program on utility-scale quantum computing. . (PsiQuantum and Microsoft are both participating in the program.
For fiscal year 2025, the Defense Department wants $69.3 million in research and development funding for quantum application in fiscal year 2025, according to budget documents. Last year’s budget request highlighted that quantum technology was reaching a “tipping point that will determine how quickly it can make an impact,” and to keep up, the U.S. needed to focus on rapidly maturing and demonstrating the tech along with the needed supply chains, according to a quantum-focused report added to the White House’s 2024 budget request.
There’s also mounting concern that China, which outspends the U.S. in key technology areas such as artificial intelligence and quantum, could access the technologies first—potentially giving that country an irrevocable advantage.
“All of the computers and all the brainpower that has ever existed are nothing compared to a very small quantum computer with just a couple of 100 logical qubits,” each of which are made up of many physical qubits that form the basic unit of quantum computing, said Denis Mandich, Qrypt’s chief technology officer, noting that the announcement could potentially change the course of history.
“With just 300 logical qubits, you have more compute power than has ever existed and could ever exist if you turned every atom—every atom in the universe—into a supercomputer,” said Mandich, who spent decades working for the CIA.
Because qubits are unstable and quickly lose their ability to run calculations, an enormous amount is needed. That is, the bigger the machine, the more physical qubits can be used to work calculations. Over time, the number of qubits needed will eventually get smaller as error correction techniques improve. On the technical side, this means the ratio of the number of physical qubits it takes to make logical qubits will keep shrinking, similar to the trajectory of the first computers.
“Right now for the superconducting quantum computers made by IBM, that number might eventually be 100 or 10,000 physical qubits to make one logical qubit. But even they’re shrinking it,” Mandich said.
But quantum isn’t just about qubits. The technology also poses real challenges to digital security—something intelligence agencies have warned of for years.
“Every message you’ve ever sent with WhatsApp or whatever is secured by just one or two algorithms…a quantum computer can break all of them, and that breaks everywhere, all at once in the entire world for every digital application,” Mandich said.
“And that’s a really big deal, because if the U.S. loses this race, the market advantage for any company that has that, especially in a country like China, is insurmountable. They win every single competition for developing new drugs for the pharmaceutical industry, for financial industry transactions, and all those things. So it’s both national security because they can read all of our secret messages, all of our chats, everything, and expose us, and they overtake huge parts of our industry with that kind of computational power.”
Right now, even with initiatives across government, the U.S. wouldn’t be ready to defend against the potential uses of quantum computers by competitors and adversaries.
For example, the Marines have nascent micro-processing and mass computation capabilities, said Lt. Gen. Matthew Glavy, the Marine Corps’ deputy commandant for information, during the Modern Day Marine conference on Tuesday.
“It’s absolutely in our best interest as warfighters to truly understand the power of that, and it is powerful,” Glavy said. “We’re still in the incline and how fast our microprocessing capabilities are based on [techniques, tactics, and procedures], based on our ability to do mass computation, supercomputing-like things. Not quantum, certainly quantum is the next step of that.”
The Navy has been working to make sure that “obsolete algorithms are out of our warfighting systems” and that “we have quantum-resistant software. So when the adversary gains that capability, we can persist against that,” Maj. Gen. Lorna Mahlock, head of the Cyber National Mission Force, said Tuesday. “And there’s a date out there we’re all working towards making sure that we’ve got the capability to be able to persist.”
The White House wants U.S. government systems to transition to quantum-resistant cryptography by 2035, according to a 2022 national security memorandum.
When asked if the U.S. would be ready if PsiQuantum or another company succeeded in building a quantum computer by 2027, Mandich said “there’s no chance that we’ll be ready. Because the last transition to new cryptography started more than 20 years ago. We’re not even done with that.”
Comments are closed.