All kinds of things are hooked up to the Internet these days, but Jerry Chow’s computer stands out. Chilled by liquid helium, his superconducting processor uses quantum physics to circumvent rules of everyday reality that limit the power of conventional computers.
Chow manages IBM’s quantum computing group at the company’s Thomas J. Watson research center in Yorktown Heights, New York. The team launched a website today with an interface that lets outside programmers and researchers test algorithms on the new chip.
Chow says he wants to get them ready for the undetermined point in the future when this exotic kind of cloud computer is ready for practical use. “We want to help people think differently and learn how to program a quantum computer,” says Chow.
IBM, Google, Microsoft, and many academic groups are all trying to develop quantum computers because they should be able to solve problems that are practically impossible for conventional computers. Not many research groups can build chips like IBM’s, and they are not usually made available to others beyond select collaborators, says David Corey, a researcher at the Institute for Quantum Computing at the University of Waterloo, Canada. And making a chip reliable enough to be available 24/7 online is impressive, he says. “I don’t know any system that shows this robustness.”
Chow’s group also published details of a second new quantum chip today. They claim it gets closer than any before to demonstrating all the error-correction features required to make a universal quantum computer, the type of quantum computer best understood by theorists. But no such machine exists because physicists haven’t figured out how to fully tame the flakiness that comes along with the potential power of representing data using delicate quantum states.
IBM’s paper is not peer reviewed, but its latest chip would put the company just ahead of Google in the race for a universal quantum computer. The search and ads giant established a new lab to build quantum chips last year (see “Google’s Quantum Dream Machine”).
Still, neither company expects to build a universal quantum computer very soon.
Quantum computing chips are made up of devices called qubits that represent digital data using quantum effects. Their computational power comes from weird tricks like entering a fragile mode called a superposition that can be thought of as simultaneously 0 and 1. A practical universal quantum computer would require hundreds of thousands or millions of qubits because of the considerable weight of error-correcting code needed. The chips IBM announced today have five and seven qubits. The best chip Google’s lead researcher has made has nine.
However, Chow’s team is also aiming at a more limited kind of quantum processor that may offer a shortcut to computational riches. A so-called analog quantum computer could operate with less error-correcting code, and hence require a much smaller number of qubits. It would only be able to take on certain problems, but they would include chemistry simulations important to energy and materials research, and machine learning, says Chow.
Scott Aaronson, an associate professor at MIT, says that’s a worthy goal. He says a collection of just 50 qubits operated that way will likely be the first computer to demonstrate “quantum supremacy”—the power to solve a computational problem immensely difficult and perhaps practically impossible for conventional machines. (Google last year reported a striking result using a quantum processor of a different design from startup D-Wave, but it fell short of quantum supremacy—see “Google Says Its Controversial Quantum Computer Really Works.”)
“The first clear demonstration of quantum supremacy will be a huge milestone in the history of physics and computer science,” says Aaronson. “It's plausible, though not certain, that it could be achieved in the near future.”
Chow says his team has that in its sights. “We’re not that far from building it,” he says. “We do believe that 50 qubits is possible in the next handful of years.”
The IBMers have competition, though. Google’s quantum computing team is also aiming to build analog quantum processors, and estimates it could have a chip with 100 qubits ready in just a few years. Academic groups, for example at the University of Maryland, are also trying to build analog quantum processors.
IBM’s Chow is unfazed, and predicts that competition will intensify as it becomes clearer what analog quantum processors might be useful for. “There are a number of companies making grabs for the IP now,” he says. “I think there will be more work on this as this scenario becomes clearer.”
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
Data analytics reveal real business value
Sophisticated analytics tools mine insights from data, optimizing operational processes across the enterprise.
Driving companywide efficiencies with AI
Advanced AI and ML capabilities revolutionize how administrative and operations tasks are done.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.