Google Quantum AI's Neven (left) and Anthony Megrant examine a cryostat refrigerator for cooling quantum computing chips at Google's Quantum AI lab in Santa Barbara, California, US. — Reuters
Alphabet Inc’s quantum computer needs just five minutes to solve a problem that would take supercomputers around 10 septillion years. Google’s next task: coming up with an actual use for all that theoretical power.
Google said its computer using the new Willow quantum chip beat the Frontier supercomputer in running a benchmark algorithm, doing in minutes what would take Frontier 10,000,000,000,000,000,000,000,000 years – far longer than the age of the universe. That’s exponentially faster than Google’s stated performance five years ago, when it said it could solve a 10,000-year task in minutes.
The algorithm, designed to test the quantum computer’s capabilities, has no known useful applications, but that’s beside the point, according to Hartmut Neven, founder of Google Quantum AI.
"If you cannot win at least on a problem, you won’t win on a useful problem either,” Neven said in an interview. Google targets delivering a real-world use case next year that would be impossible for a classical computer to solve, he said. "That is now getting within reach.”
Governments, as well as some of the world’s biggest tech companies and venture capitalists, have poured billions of dollars into quantum computers, lured by the promise of commercial and military supremacy arising from computing speeds millions of times greater than that of classical computers.
But because quantum computers harness the behavior of subatomic particles, they need to operate in environments that prevent the particles from interacting with their surroundings, with most experiments assuming temperatures near absolute zero. Such limitations have made it difficult to find practical and real-world applications for the technology, as high error rates made it difficult to conduct quantum computing at scale.
The Willow chip lowers error rates, according to a paper published in the scientific journal Nature on Monday. That makes it possible to build a bigger quantum computer, and Google’s now in a position to begin weighing costs, Neven said.
Various methodologies are vying to win the quantum computing arena. Google’s technology is called superconducting qubits, also used by rivals International Business Machines Corp and Amazon.com Inc, and the Willow chips are made with tools similar to those used to make conventional microchips. But Google also recently invested in QuEra Computing Inc which uses what’s known as neutral atom qubits. A qubit is the basic unit of information in quantum computing.
"When we make that decision to pull the trigger to scale up, we want to be absolutely certain we scale up the most promising technology. Our money is on that this would be superconducting qubits,” said Neven. "But maybe QuEra teaches us that neutral atoms have their advantages. We’ll see.” – Bloomberg
