Quantum computers are already here, even though it’s not readily apparent. Now, researchers say quantum advantage—the field’s long-promised milestone of outperforming classical computers—appears to have finally arrived. But the story comes with an important caveat.
Research by scientists at the University of Texas at Austin and Colorado computing firm Quantinuum devised and carried out an experiment that demonstrates “unconditional” quantum advantage, sometimes referred to as quantum supremacy. As the researchers phrased it, their “result is provable and permanent: no future development in classical algorithms can close this gap.” The preprint, which has yet to be peer reviewed, was made available on arXiv earlier this month.
Gizmodo reached out to several experts in the field, who affirmed the new results. They added that the experiment, while commendable, isn’t the most practical use of a quantum computer—which already gets flak for its uselessness to everyday users.
Then again, “quantum advantage” is a weird, surprisingly malleable concept with many possible applications. Overall, the results are definitely worth a closer look.
Alice and Bob make a cameo
Quantum enthusiasts may be familiar with Alice and Bob, two fictional characters often summoned for quantum thought experiments. In the context of the new experiment, Alice and Bob are two researchers collaborating on a computation using a single device. They receive different inputs at different points in time, but only Alice can send Bob a message, and not the other way around. Based on Alice’s message, Bob must decide how to measure and interpret to produce a final output.
According to the paper, “the use of a quantum message can provably reduce the amount of communication required by an exponential factor compared to any protocol that uses classical communication alone.” In other words, a small quantum message can replace a much larger classical one. To prove their point, the team repeated the experiment 10,000 times on Quantinuum’s H1-1 trapped-ion quantum computers, coupled with a careful mathematical validation of their protocol.
Surprisingly, they found that a quantum computer only needed 12 qubits (qubits are the smallest unit of information for quantum computers) to solve this problem. By contrast, even the most efficient classical computers needed 330 bits.
A different way to play the game
“This is a very different type of quantum advantage than we have seen before—not better or worse, but it’s just proving something completely different from past experiments,” Bill Fefferman, a computer scientist at the University of Chicago, told Gizmodo in an email. Fefferman previously collaborated with senior author Scott Aaronson but wasn’t involved in the new study.
Fefferman explained that scientists typically equate quantum advantage to “striving to perform a computation on a quantum computer that can be solved dramatically faster than any classical computer.” By contrast, the new experiment achieves “quantum information supremacy,” in which the focus isn’t so much on speed as it is on using fewer qubits to solve a problem that classical computers need many more bits to crack.
“It is true that their result is unconditional, in the sense that it doesn’t rely on unproven assumptions,” Fefferman said. “This is, of course, a great feature of this new experiment, but it’s also inherited by this ‘moving of the goalposts.’”
Gizmodo contacted the study’s authors, who said they couldn’t comment until the paper is formally published.
Pressing the advantage
The results raise questions about the broader goals of proving quantum advantage. As IBM Quantum’s director told Gizmodo in a previous interview, a potential answer is to ask how quantum computers can enhance computing problems we’re already familiar with.
But as Fefferman noted, there isn’t necessarily a better or worse approach for arriving at quantum advantage—although this “goalpost” appears to be the holy grail for the field’s struggle to prove its worth.
That may be a product of quantum computing’s history, Giuseppe Carleo, a computational physicist at EPFL in Switzerland who wasn’t involved in the new work, explained to Gizmodo in a video call. The rapid growth of quantum computing makes it easy to forget how recently the right hardware became available to test theory.
“So the field has developed historically in the past 20, 30 years much closer to mathematics, rather than an applied field where, if you want, you can use a machine to run things,” said Carleo, who spoke with Gizmodo about the history of quantum computing. As a result, most of the analysis in the field remained at theoretical levels for a longer time than scientists would’ve hoped.
But with hardware advances and a fast-growing industry, this trend is gradually shifting—as it should, Carleo said. More projects are moving away from designing quantum advantage experiments “specifically tailored to show advantage,” he said, turning instead to places where quantum computers can help, not necessarily upend.
That’s actually closer to the field’s “origins,” he added. Richard Feynman, the physicist instrumental to quantum computing’s foundations, suggested that quantum computers should predict quantum phenomena. Sure, there might not be so much “money attached to it,” but they are “of tremendous interest for theoretical physics,” particularly with regard to fundamental questions about our universe, Carleo explained.
Quantum-anything never makes it easy
The new experiment might struggle to prove its immediate connection to practicality. But in a way, the preprint does adhere to Feynman’s advice. It’s certainly a theoretically robust demonstration of using quantum hardware to investigate quantum concepts.
At this very moment, that makes it seem detached from reality. Then again, when has anything quantum ever given easy answers? Yet, if science history is any guide, the best discoveries come from the most unexpected, seemingly impractical pursuits. We’ll just have to keep watch.
Read the full article here
