Google presented its new quantum chip. The company claims that its computing power is so powerful that in 5 minutes it can solve a problem that would take the currently fastest supercomputer ten septillions, or 10,000,000,000,000,000,000,000,000 years.
The Google chip, dubbed Willow, is the latest development in a field called quantum computing, which attempts to use the principles of particle physics to create a new type of computer with enormous power.
Google says its new quantum chip contains key “breakthroughs” and “paves the way to a useful quantum computer at scale.” The company claims that it has such computing power that it can solve in 5 minutes a task that would take the currently fastest supercomputer ten septillion years. That is, 10,000,000,000,000,000,000,000,000 years.
However, as the BBC points out, experts say Willow is a largely experimental device for now, which means it will still take years – and billions of dollars – to create a quantum computer powerful enough to solve a wide range of real-world problems.
Quantum computers
Quantum computers work in a fundamentally different way than the computer in your phone or laptop. They use the principles of quantum mechanics, a field of science devoted to the behavior of ultrafine particles, to solve problems much faster than traditional computers.
Scientists hope that quantum computers will eventually be able to use this ability to significantly speed up complex processes such as creating new drugs. There are also concerns that such a computer could be used for malicious purposes – for example, to break some types of encryption used to protect sensitive data.
In February, Apple announced that the encryption protecting iMessage chats is “quantum-resistant” to prevent them from being read by future powerful quantum computers.
Hartmut Neven, who heads the Google Quantum AI lab that created Willow, describes himself as the “main optimist of the project.” He told the BBC that Willow would be used in some practical applications, but declined to provide further details for now. – However, a chip capable of commercial applications will only appear at the end of the decade – he added.
Initially, these applications would concern the simulation of systems in which quantum effects are important.
– For example, they are important for the design of fusion reactors, as well as for understanding the functioning of drugs and pharmaceutical development. This would also be important for developing better car batteries and much more, Neven said.
Scientists' cautious skepticism
Neven told the BBC that Willow's performance means it is “the best quantum processor ever built”.
However, Professor Alan Woodward, an expert in computer science at the University of Surrey, in an interview with the British website, points out that quantum computers will be better at many tasks than current “classical” computers, but will not replace them. He cautions against overestimating the importance of Willow's achievements based on her scores on a single test.
“You have to be careful not to compare apples and oranges,” he told the BBC.
Google chose a problem to use as a performance benchmark that was “tailor-made for a quantum computer” and did not demonstrate “universal speedup compared to classical computers.” Nevertheless, he said Willow represents significant progress, particularly in so-called error correction.
Simply put, the more useful a quantum computer is, the more qubits it has. However, the technology's main problem is its error-proneness – a tendency that previously increased the more qubits a chip had.
However, Google researchers say they have managed to reverse this and design and program the new chip so that the error rate decreases throughout the system as the number of qubits increases. It was a major “breakthrough” that addressed a key challenge that had faced the field “for almost 30 years,” Neven believes.
He told the BBC it could be compared to having “a plane with just one engine – that will work, but two engines are safer, four engines are even safer.” Errors are a significant obstacle to creating more efficient quantum computers, and the development has “encouraged everyone who has tried to build a practical quantum computer,” Professor Woodward said.
But Google itself notes that to develop practically useful quantum computers, error rates will still need to be much lower than those found with Willow. Willow was manufactured at Google's new, purpose-built manufacturing facility in California.
Not a breakthrough, but an impressive work
Michael Cuthbert, director of the UK's recently established National Quantum Computing Center (NQCC), told the BBC that he was concerned about using language that fuels excitement and hype around a topic. He considers Willow to be more of a “milestone than a breakthrough”, but nevertheless stated that it is “clearly a very impressive work”.
Ultimately, quantum computers would help with a range of tasks, including “logistical problems such as distributing cargo on planes or routing telecommunications signals or storing energy across a national grid,” he said.
There were already 50 quantum companies in the UK, with funding worth £800 million and employing 1,300 people.
On Friday, scientists from the University of Oxford and the University of Osaka in Japan published a paper showing very low error rates in a qubit with trapped ions. Their approach differs by creating a quantum computer capable of operating at room temperature, while Google's chip must be stored at very low temperatures to be effective.
Main photo source: Gorodenkoff/Shutterstock