Alphabet Inc.’s Google ran an algorithm on its “Willow” quantum-computing chip that can be repeated on similar platforms and outperform classical supercomputers, a breakthrough it said clears a path for useful applications of quantum technology within five years.
The “Quantum Echoes” algorithm, detailed in a paper published Wednesday in the science journal Nature, is verifiable, meaning it can be repeated on another quantum computer. It also ran 13,000 times faster than possible on the world’s best supercomputer, Google said. Taken together, the advances point to a broad range of potential uses in medicine and materials science, Google said.
“The key thing about verifiability is it’s a huge step in the path toward a real world application,” said Tom O’Brien, a staff research scientist at Google Quantum AI who oversaw the completion of this work. “In achieving this result we’re really pushing us toward finding mainstream.”
The breakthrough brings Google a step closer to harnessing the processing power promised by quantum computing, also being pursued by rivals Microsoft Corp., International Business Machines Corp. and numerous startups. It follows Google’s announcement in December that Willow had solved a problem in five minutes that would have taken a supercomputer 10 septillion years.
Quantum computers use tiny circuits to perform calculations, like traditional computers do, but they make these calculations in parallel, rather than sequentially, making them much faster. While firms have boasted of building quantum platforms that surpass classical computers, their challenge has been to find a useful application.
Computer scientist Scott Aaronson, who wasn’t involved in the study, wrote in an email that he was “thrilled” by Google’s progress toward outperforming supercomputers in a way which could be efficiently repeated, and thus proved, on a second quantum computer — which had been “one of the biggest challenges of the field for the past several years.” Still, he warned that there was a lot of work ahead.
“Getting from here to anything commercially useful, and/or to scalable fault-tolerance (which wasn’t used for this demonstration), will be additional big challenges,” wrote Aaronson, who serves as the Schlumberger Centennial Chair of computer science at the University of Texas at Austin.
One use of the algorithm is to examine molecular structures by computing the distances between atoms, scientists showed in a collaborative second paper that has not yet been peer-reviewed. The method could be applied to drug discovery and material science, including battery design, though that would require a quantum computer 10,000 times larger than current working machines, Google scientists estimated.
The Google team, which includes 2025 Nobel Prize in Physics winner Michel H. Devoret, said it plans to continue to move toward real-world applications by scaling up and improving the accuracy of its machines.
. Read more on Technology by NDTV Profit.