Quantum computers promise to solve problems far beyond the reach of today’s most powerful supercomputers, but making them practical has been a challenge.
Now, a breakthrough by UNSW Sydney spinout Diraq suggests that the path to building large-scale, cost-effective quantum computers may finally be taking shape.
Diraq specializes in silicon-based quantum chips—devices that harness the rules of quantum mechanics to perform calculations.
Until now, these chips had shown high performance only in controlled laboratory environments.
The big question was whether they could deliver the same results when produced in industrial settings, using the same methods that create everyday computer chips.
That question has now been answered.
In collaboration with Europe’s Interuniversity Microelectronics Center (imec), one of the world’s leading semiconductor research institutes, Diraq has proven that its chips can achieve the 99% accuracy required for reliable quantum operations, even when manufactured outside the lab.
“This is a crucial step,” said Professor Andrew Dzurak, UNSW engineer, founder, and CEO of Diraq. “For the first time, we’ve shown that the fidelity—the accuracy of our quantum processors—can be reproduced in a commercial semiconductor foundry.
That means our chips are fully compatible with the manufacturing processes that have been perfected over decades.”
The results, published in Nature, focus on the performance of operations involving two quantum bits, or qubits. Two-qubit “logic gates” are the fundamental building blocks of quantum computers, and demonstrating that they can be fabricated with high accuracy in an industrial environment is a major milestone.
The achievement also aligns with the Quantum Benchmarking Initiative, a program led by the U.S. Defense Advanced Research Projects Agency (DARPA). Its goal is to determine when quantum computers will reach “utility scale”—the point at which their commercial value outweighs their cost of operation.
To reach that threshold, quantum processors will eventually need to manage and correct errors across millions of qubits.
Silicon is emerging as the leading candidate for achieving this. Unlike some exotic materials being tested elsewhere, silicon can pack millions of qubits onto a single chip and is already the backbone of the global microchip industry. This makes it far easier to scale up and much more cost-effective.
Diraq had previously shown that single-qubit operations could reach 99.9% accuracy using standard CMOS (complementary metal-oxide-semiconductor) processes, the same technology behind ordinary computer chips. But until now, demonstrating equally strong results for two-qubit operations in a factory environment remained out of reach.
The new results change that. “This latest achievement clears the way for building a fully fault-tolerant, functional quantum computer that is more cost-effective than any other qubit platform,” said Dzurak.
By proving that high-fidelity quantum chips can be manufactured at scale, Diraq and its partners may have just taken the most important step yet toward making quantum computers a reality.
