Google, IBM and the Quantum Milestone That Keeps Moving

The word "milestone" has done a lot of work in quantum computing. Google's claim of quantum supremacy, published in Nature's 150th anniversary issue in October 2019, is the loudest one. It is not the only one, and it is not the one that ends the argument.

The claim itself is narrow and specific. According to Google's own blog post, its Sycamore machine performed a test computation in 200 seconds that the best known algorithms on the most powerful supercomputers would need thousands of years to replicate. In the paper, reported by Fortune on 20 September 2019, the researchers put a finer point on it. The processor took about 200 seconds to sample one instance of the quantum circuit a million times. A state-of-the-art supercomputer would need approximately 10,000 years for the equivalent task. The same paper estimated that running the experiment on a Google Cloud server would take 50 trillion hours.

Fortune obtained a copy of the paper after it appeared briefly on NASA.gov and was taken down. The Financial Times first reported the news. A Google spokesperson declined to confirm the paper's authenticity. A source at Google suggested NASA had published it early, before peer review, a process that could take weeks or months.

Two years of claims, one disputed word

IBM got there first with a number, if not with the same word. On 10 November 2017, the Associated Press reported that IBM scientists had built and measured a processor prototype with 50 quantum bits. Dario Gil, who led IBM's quantum computing and artificial intelligence research division, said it was the first time any company had built a quantum computer at that scale. Seth Lloyd, an MIT mechanical engineering professor not involved in the work, said IBM likely still had glitches to work out. He called the announcement a sign of significant progress.

Google's own path to that scale ran through smaller chips. In April 2017, MIT Technology Review reported that John Martinis, who leads Google's quantum research group, had given himself until the end of that year to reach quantum supremacy. The team had a nine-qubit chip arranged in a line, and its newest chip had six qubits in a two-by-three configuration. Martinis said he needed a grid of 49 qubits for the experiment. Designs for 30 to 50 qubits were already in progress.

By December 2017, Martinis was telling a conference at NASA Ames that qubit count was the wrong headline. "Press releases always talk about quantum space race in number of qubits," he said, according to Gizmodo. "It is about quantity, but also qubit quality." Google was then fabricating its 49 or 50 qubit device, with testing to begin within two weeks.

The final machine was a 53-qubit processor called Sycamore, scaled back from a 72-qubit design named Bristlecone. Google's blog post framed the result as the "hello world" moment the field had been waiting for. It also conceded that it will be many years before anyone can implement a broader set of real-world applications.

What the milestone does not measure

IBM's Dario Gil told Fortune that the experiment was a highly special case laboratory experiment with no practical applications, and he advised against using quantum supremacy as a progress metric. "The experiment and the 'supremacy' term will be misunderstood by nearly all," he said. "Quantum computers will never reign 'supreme' over classical computers, but will rather work in concert with them, since each have their unique strengths."

Jim Clarke, Intel Labs' director of quantum hardware, called the update a notable mile marker. He said a commercially viable quantum computer would require many research and development advances first. "While development is still at mile one of this marathon, we strongly believe in the potential of this technology," he said.

That caution has a history. Chris Monroe, a University of Maryland professor and cofounder of the startup IonQ, told MIT Technology Review in 2017 that supremacy would be "an academic milestone." "Afterward you still have to figure out how to make it more scalable and programmable," he said.

IBM's own metric work points the same way. At the American Physical Society March Meeting in 2019, IBM outlined a Quantum Volume of 16 for its 20-qubit Q System One, double the Quantum Volume of 8 for its then-current IBM Q system. Quantum Volume accounts for qubit count, connectivity and coherence time, plus gate and measurement errors, device cross talk and circuit software compiler efficiency. IBM said Quantum Volume would need to double every year to reach Quantum Advantage within the next decade. That is the point where quantum applications deliver significant advantages over classical computers. IBM has doubled the power of its quantum computers annually since 2017.

The Google paper's own conclusion is more ambitious than the experiment. The researchers predicted that quantum computing power will grow at a double exponential rate, outpacing the exponential rate that defined Moore's Law. They also wrote: "We are only one creative algorithm away from valuable near-term applications."

That is the part still untested. The 200-second run is a data point. The 10,000-year comparison is a data point. Neither tells a company what to do on Monday.