Word that a Google-led team of researchers had achieved “quantum supremacy” with a new type of computer chip leaked out weeks ago, but today’s publication of the team’s study in the journal Nature gave outsiders their first good look at what was done. And most of them were impressed.
There were the usual caveats, of course: The project focused on a specific problem in random number generation that’s doesn’t relate directly to everyday applications, and it could be years before the technology behind Google AI Quantum’s Sycamore chip becomes commercially available.
Nevertheless, the computational demonstration provided evidence that quantum computers can do some tasks far more quickly than classical computers.
“This is an exciting scientific achievement for the quantum industry, and another step on a long journey towards a scalable, viable quantum future,” Microsoft, one of Google’s competitors in the realm of quantum computing, said in a statement emailed to GeekWire.
In February, Microsoft formed its own network for quantum startups and developers, and just weeks later the company played a prominent role alongside the University of Washington and the Pacific Northwest National Laboratory in organizing the Northwest Quantum Nexus.
“For the last two years, we have been working to deliver quantum impact for our customers today,” Microsoft said. “Customers like Case Western Reserve University are seeing real benefits with our quantum-inspired algorithms, and nearly 200K developers have contributed to the growing quantum economy through our open-sourced Quantum Development Kit.”
Quantum information science is one of the technological priorities for the federal government. Today’s research revelations sparked a glowing op-ed in Fortune from White House chief technology officer Michael Kratsios, who said the “giant leap forward in quantum computing” came about due to “America’s unparalleled innovation ecosystem.”
Even Ivanka Trump, President Donald Trump’s daughter and a White House adviser, got into the act. “The US has achieved quantum supremacy!” she tweeted.
Why all the fuss?
Quantum computing is one of the farthest-out frontiers in computer science, and promises to facilitate breakthroughs in cryptography, molecular chemistry, battery materials and high-temperature superconductors.
Classical computers are based on a foundation of zeroes and ones. In contrast, quantum computing devices make use of quantum bits, or qubits, that can hold multiple values simultaneously during processing. In accordance with the spooky world of quantum physics, the results collapse into definite values only when they’re read out.
D-Wave Systems, based near Vancouver, B.C., has developed devices that use a technique known as quantum annealing for limited computational purposes. Other companies, including Google, Microsoft and IBM, are trying to go further by creating general-purpose quantum computers.
Not every type of computational problem is suited for the quantum approach — but computer scientists have claimed that quantum computers should theoretically be able to solve some types of problems far more easily than classical computers. The experiment described in the paper published today suggests that such claims are no longer merely theoretical.
Researchers made use of a 54-qubit Sycamore processing chip that incorporates fast, high-fidelity quantum logic gates. The system was designed to reduce error rates, which is one of the chief bugaboos for an interconnected network of qubits.
The system was tested by having it produce progressively longer sequences of numbers, corresponding to the configuration of random quantum circuits.
“Our machine performed the target computation in 200 seconds, and from measurements in our experiment we determined that it would take the world’s fastest supercomputer 10,000 years to produce a similar output,” lead Google researchers John Martinis and Sergio Boixo reported today in a blog posting.
They said the experiment successfully challenged the hypothesis that classical computers can efficiently implement any reasonable model of computation. That’s what’s meant by the term “quantum supremacy.”
“With the first quantum computation that cannot reasonably be emulated on a classical computer, we have opened up a new realm of computing to be explored,” the researchers wrote.
IBM cast doubt on Google’s claim, saying that its Summit supercomputer could have performed the task with far greater fidelity in far less than 10,000 years — 2.5 days, to be precise — but today Google cast doubt on IBM’s doubts.
The Sycamore quantum computer is fully programmable and can run general-purpose quantum algorithms, which means there’ll be further tests ahead.
“Our team has already been working on near-term applications, including quantum physics simulation and quantum chemistry, as well as new applications in generative machine learning, among other areas,” Martinis and Boixo said.
They said they’re continuing efforts to make the Sycamore system more fault-tolerant, and will make the processors available to collaborators and academic researchers for further experiments. “Creative researchers are the most important resource for innovation — now that we have a new computational resource, we hope more researchers will enter the field motivated by trying to invent something useful,” Martinis and Boixo said.
Martinis and Boixo are among 77 authors of the Nature paper, titled “Quantum Supremacy Using a Programmable Superconducting Processor.” The authors come from institutions including Google AI Quantum as well as the Quantum AI Laboratory (QuAIL) at NASA’s Ames Research Center, the University of California at Santa Barbara, Oak Ridge National Laboratory and Forschungszentrum Jülich.