IBM has a roadmap to a ‘fault-tolerant' Quantum computer by 2029

Mint

2 days ago

International Business Machines said Tuesday it has a plan for building what it calls the world's first large-scale, fault-tolerant quantum computer at its New York data center before the end of the decade.
The computer, called IBM Quantum Starling, will be housed in its Poughkeepsie, N.Y., center and have 20,000 times the computational power of today's quantum computers, the tech giant said.
'I feel more comfortable than ever that a fault-tolerant quantum computer will exist before the end of this decade," said Jay Gambetta, IBM's vice president of quantum. 'We are putting error-correction in detail on our roadmap because we believe now we've solved all the scientific challenges."
Quantum computers are susceptible to instability, requiring quantum error correction—a technique that identifies and addresses errors in computations—and more broadly, quantum fault-tolerance, the ability to operate even with errors present.
Among the companies racing to build a practical quantum computer, from tech giants like Microsoft and Google to quantum companies like D-Wave, Quantinuum and IonQ, IBM isn't the only one that has promised a fault-tolerant computer by 2029.
Others have recently made progress in the area of error correction, including Amazon, with its Ocelot quantum computing chip, which it said can reduce quantum errors by up to 90%. Google is also focused on error correction with its Willow chip.
Quantum computers store information as quantum bits—otherwise known as qubits—which can exist as a zero and a one at the same time and so are much richer objects than the binary digits that ordinary computers use. That makes them capable of much more powerful types of computations than ordinary computers, and could mean they can help engineer materials at the molecular level, or even crack the defenses used to secure the internet.
But one of quantum computing's major problems is that qubits generate errors as they tackle problems. They're fragile, and susceptible to 'noise," essentially small environmental disturbances that can force them out of their quantum state.
That's made the quest to build a so-called fault-tolerant quantum computer a priority for tech giants and quantum companies alike.
Gambetta said IBM's confidence in its 2029 timeline stems from two recent developments: further advances in a new approach to reducing errors called 'quantum low-density parity check" or qLDPC codes, plus a technique for identifying and correcting errors in real-time using conventional computing.
Additionally, quantum computing startup SEEQC is expected to announce a collaboration with IBM this week as part of the U.S. Defense Advanced Research Projects Agency's Quantum Benchmarking Initiative, which is aimed at evaluating quantum companies' ability to achieve utility-scale operation.
As part of the partnership, the two companies will work to shrink and consolidate IBM's racks of control hardware by transferring some control and readout elements directly onto a chip. SEEQC manufactures its own quantum chips, known as integrated QPUs, in a factory in Elmsford, N.Y.
IBM is releasing more detailed plans for its large-scale quantum computer at this point because it hopes to spur developers' interest in creating quantum algorithms—a key part of the software that will run on quantum computers. That's critical to achieving a return on investment for quantum computers, a metric that matters to businesses, Gambetta said.
But, at this point, it's not clear how IBM's breakthroughs 'translate into tangible business value," and the transformative potential of fault-tolerant quantum computers is still speculative, said Chirag Dekate, an analyst at IT research and consulting firm Gartner.
IBM's plans are also light on details about its new quantum computer's commercial availability, and specific dates for when its error-corrected systems will be released, Dekate said. In response, IBM asserted that it has the most detailed roadmap and plans for commercialization in the industry.
'This is inherently hard to do because quantum innovation has many moving parts," Dekate said. 'The reality in quantum is that we are not yet at the ChatGPT-like moment where the technology, algorithms and impact become visceral and undeniable."
At the same time, the comprehensive nature of IBM's plans to build its fault-tolerant quantum computer is worthy of note, said Heather West, an analyst at International Data Corp.
IBM's plans should serve as a wake-up call to businesses and the tech community to pay attention to the speed at which quantum computing is advancing, West added.
'When you hear this announcement by IBM, as well as others by other hardware vendors, you realize that this is a real technology, and it's coming to fruition as we speak," she said.
Write to Belle Lin at [e-mail] and Isabelle Bousquette at [e-mail]