International Business Machines Corp. has signed up several prominent banks, as well as industrial and technology companies, to start experimenting with its quantum computers.
JPMorgan Chase & Co. and Barclays Bank Plc are among the 12 organizations IBM selected to get early access to its latest quantum computing technology, which could eventually help solve difficult problems that are beyond the reach of any conventional supercomputer.
Automakers Daimler AG and Honda Motor Co., technology giant Samsung Electronics Co., chemicals companies JSR Corp. and Nagase & Co., and specialty materials company Hitachi Metals Ltd. are also in the group.
Many of the most promising potential applications for quantum computers involve creating new chemical catalysts or engineering new materials with exotic properties.
IBM is competing with Alphabet Inc.’s Google, Microsoft Corp., Intel Corp., Canadian company D-Wave Systems Inc. and California-based Rigetti Computing, as well as a number of other small startups to commercialize the technology. Many of these companies plan to offer access to quantum computers through their cloud computing networks and see it as a future selling point. For now, quantum computers still remain too small and the error rates in calculations are too high for the machines to be useful for most real-world applications.
IBM will give the companies it has selected for its early access program — called the Q Network — access to a 20-qubit quantum computer through the cloud. While such a computer can theoretically perform more than 1 million simultaneous calculations, that compares to classically designed supercomputers that can perform tens of thousands of trillions of calculations per second. What’s more, quantum computers like IBM’s can’t make calculations without accruing errors that can make its results useless for many practical applications. By giving the companies access to quantum computers, however, IBM is banking on hooking them early on the emerging technology so that when more powerful machines are ready the companies will already be comfortable with the systems.
IBM has also recently tested a prototype of a 50-qubit quantum computer — which is approaching a threshold known as “quantum supremacy,” where it might be able to perform calculations that are beyond the reach of a classical supercomputer. But it still produces errors. IBM said it would let the companies in its early access program begin experimenting with this size machine through the cloud in the near future.
Daimler said it would investigate how quantum computers might help it perfect its manufacturing processes, route fleets of self-driving cars or help the company develop new materials for automobiles. Samsung said it would be looking at what impact quantum computing might have on semiconductors, while JSR said it would research chemical and material applications. JPMorgan said it would partner with IBM to explore whether quantum computers can be used for trading strategies, portfolio optimization, asset pricing and risk analysis.
Classical computers work using tiny transistors imprinted on silicon chips. They process information in a binary format, either a 0 or 1, called bits. Instead of bits, quantum computers use qubits, which can represent both a 0 and 1 simultaneously and give quantum computers much more power than classical computers. But they are prone to errors and must be cooled to temperatures below those found in outer space. Microsoft is working on a different type of qubit that will theoretically produce a lower error rate, making it more useful for commercial applications. But the company has not yet managed to build a working model.
IBM and the other companies in the race to commercialize the technology, however, have begun offering customers simulators that demonstrate what a quantum computer might be able to do without errors. This enables companies to begin thinking about how they will design applications for these machines. In some cases, like machine learning, quantum-inspired algorithms have helped companies improve the efficiency of the software they run on standard computers.
Other participants in the IBM’s Q Network include the Oak Ridge National Laboratory in the U.S., Keio University in Japan, the University of Oxford, in the U.K., and the University of Melbourne, in Australia.
By Jeremy Kahn