The chip, dubbed Ocelot, can lower the costs of reducing quantum computing errors by up to 90%, the Seattle-based tech giant said.
The race to build a practical quantum computer has been heating up among startups and the world’s biggest tech companies. Amazon Web Services’ announcement comes just a week after Microsoft claimed a quantum computing breakthrough by creating a new state of matter. Google in December said it developed a new chip, called Willow, that it said marks an advance in solving quantum’s error-correction issues.
Amazon’s Ocelot chip is a prototype, not a “full-blown quantum system,” said Oskar Painter, head of quantum hardware for AWS. “It’s designed to test our ability to perform quantum error correction, and once we have that building block, then we can scale it up to a much larger size.”
One of quantum computing’s biggest challenges is that qubits, the basic units of quantum computation, generate errors as they tackle problems, with the errors increasing as quantum systems get bigger. Qubits are fragile and susceptible to “noise,” essentially small disturbances in the environment like heat and vibrations, which force them out of their quantum state. Quantum error correction is used to help fix the problems created from that noise.
Quantum computers can crunch numbers in a fundamentally different way from traditional computers, and can do certain computations orders of magnitude faster. They can be useful in areas including drug discovery, data encryption and cybersecurity.
While some companies have been working on building practical quantum computers for decades, the technology is still at least a decade away from commercial use, according to Painter. Other companies and industry experts say it’s just a few years away.
Amazon, Microsoft and Google’s approaches are all trying to solve the challenge of quantum error correction, though Google’s work in developing superconducting quantum circuits is most similar to Amazon’s, according to Painter. Results of Amazon’s work were published in Nature, a leading scientific journal, on Wednesday.
Amazon formally kicked off its efforts to build a quantum computer about five years ago with the creation of a quantum lab on the California Institute of Technology’s campus in Pasadena, Calif.
That decision was made to purposefully straddle the lines between academia and industry, Painter said. Since the lab’s founding, the team—which is also located in San Francisco, Boston, and New York—has grown to over a hundred, he said.
Amazon has created a new approach for quantum error correction, said Paul Smith-Goodson, a quantum computing analyst at technology research firm Moor Insights & Strategy. And though it entered the quantum race later, it could still catch up to competitors if its method is proven out, he said.
Still, there’s a fair amount of skepticism that should be applied to any quantum development, said Heather West, an analyst focused on quantum computing at research firm International Data Corp.
“I would categorize this as much more of an advancement and less of a breakthrough,” West said, referring to Amazon’s Ocelot chip. The company’s approach also isn’t completely novel, West said, because others have already demonstrated the usefulness of superconducting qubits designed to be more resistant to certain types of errors.
“There’s not at this point a known solution to how we’re going to reduce these errors in quantum systems,” she added.
For Amazon, the launch of its first quantum chip is part of a larger effort to get businesses on board with quantum computing. Last year, the company launched a quantum-computing advisory program that aims to prepare customers for society’s shift toward the technology. In 2019, it announced a cloud-based service called Amazon Braket, which lets businesses develop and test quantum algorithms in simulations.
Amazon’s next quantum update is likely a few years away as it continues building up what Ocelot is capable of, Painter said.
“We’re starting to get closer to a tipping point where these are going to become real technologies,” he said.
Write to Belle Lin at belle.lin@wsj.com.