Honeywell’s second-generation quantum computer, the H1, is in business. The powerful computer performs calculations by carefully manipulating 10 ytterbium atoms housed in a thumbnail-size package called an ion trap.
Honeywell, a surprise new entrant into quantum computers, is one of a several companies hoping to revolutionize computing. Tech giants IBM, Google, Intel and Microsoft also have serious quantum computing programs, and startups such as Rigetti Computing and IonQ are in the fray with their own machines.
A host of other startups like QC Ware, Zapata, Cambridge Quantum Computing, Rahko, and Xanadu are working to make quantum computers easier to use for those that don’t have a bunch of Ph.D.s on staff to wrestle with the weird laws that govern the ultra-small scale of the quantum physics realm.
The continued progress is essential if quantum computers, still in their infancy, are to meet their potential. Years of investments will be required to carry today’s early designs to a more practical, profitable phase.
The heart of a quantum computer is called a qubit, a data storage and processing element that unlike conventional computer bits can store an overlapping combination of zero and one through one quantum computing phenomenon called superposition. Honeywell’s H1 machine today has 10 qubits, charged ytterbium atoms arranged in a line.
Tickling qubits to perform quantum calculations
Those qubits can be tickled electromagnetically to change the data they’re storing, shift positions and reveal their state to the outside world when a calculation is finished. Qubits can be connected through a phenomenon called entanglement that exponentially increases the number of states a quantum computer can evaluate.
That’s why quantum computers promise to be able to crack computing problems that conventional machines can’t. One big expected use is molecular modeling to improve chemical processes like fertilizer manufacturing. Quantum computers are also expected to take on other materials science challenges, such as creating efficient solar panels and better batteries. Other uses focus on optimization tasks like overseeing the financial investments and routing a fleet of delivery trucks.
Honeywell pioneered this trapped-ion design with the H0 quantum computer prototype. “Because of demand from partners and customers, we transformed H0 into a commercial system,” said Tony Uttley, president of Honeywell Quantum Solutions. Customers who’ve used H0 include Los Alamos National Laboratory and the University of Texas at Austin, oil-and-gas giant BP and financial services company JPMorgan Chase.
Honeywell H0 today to H5 in 2030
The H0 set a record for an IBM-designed quantum computing speed test called quantum volume, a measure that combines the number of qubits and how much useful work they can accomplish. In August, IBM reached a quantum volume of 64, part of a plan to double performance annually. But in October, Honeywell announced its H0 reached a quantum volume of 128. That’s part of its plan to increase performance at least by a factor of 10 annually, reaching 640,000 by 2025.
Honeywell also detailed H2, H3, H4 and H5 quantum computer design plans extending through 2030. They’ll replace today’s straight-line ion trap with increasingly complicated arrangements, including a looped “racetrack” in the H2 already in testing today and increasingly large crisscrossing lattices for the H3, H4 and H5.
One big motivation for the new designs is cramming in more qubits. That’ll be important to move beyond today’s kicking-the-tires calculations into more serious work. It’ll be essential for one of the big challenges for future quantum computers, error correction, which designers hope will let easily perturbed qubits perform calculations for longer before being derailed.