Microsoft's

Microsoft is planning to build a quantum supercomputer within 10 years

Microsoft today outlined its strategy for developing its quantum supercomputer, which would use topological qubits developed by the company's researchers for several years. There are still many intermediate milestones to be met, but Microsoft's VP of advanced quantum development, Krysta Svore, told us that the company believes it will take less than ten years to build a quantum supercomputer using these qubits that can reliably perform one million quantum operations per second. That's a new metric introduced by Microsoft as the industry strives to advance beyond the present age of noisy intermediate-scale quantum (NISQ) computing.

"We think about our roadmap and the time to the quantum supercomputer in terms of years rather than decades," Svore stated.

Microsoft reported a significant advance last year when its team first demonstrated its ability to manufacture Majorana-based qubits. Majorana qubits are exceedingly stable (particularly when compared to typical procedures), but they are also extremely difficult to construct. Microsoft bet early on this technology, and now, a year after first announcing this milestone, the team is publishing a new peer-reviewed paper (in the American Physical Society's Physical Review B) proving that it has indeed achieved this first milestone on its way to a quantum supercomputer.

To get to this stage, Microsoft displayed findings from significantly more devices and data than when it initially disclosed this study a year ago. "Today, we're really at this foundational implementation level," Svore stated. We have a lot of noise in our intermediate-scale quantum devices. They're based on physical qubits and aren't yet dependable enough to perform anything practical or advantageous in terms of utility. For scientific research or commercial purposes. The robust level is the next one we need to reach as an industry. We must be able to work not just with physical qubits, but also to incorporate those physical qubits into an error-correcting code and use them as a unit to act as a logical qubit."

Svore contends that reaching this position will necessitate a quantum computer capable of doing one million trustworthy quantum operations per second with a failure rate of one trillion operations.

The next stage is to create hardware-protected qubits, and Svore stated that the team is making significant progress in this regard. These qubits will be tiny (less than 10 microns on a side) and fast enough to conduct a single qubit operation in under a microsecond. Following that, the team intends to work on entangling these qubits and operating them via a technique known as braiding, which has been explored (mainly as a theory) since at least the early 2000s.

Following that, a smaller multiqubit system will be built and a full quantum system will be demonstrated.

That is an ambitious vision, and given how long Microsoft took to reach even the first milestone, we will have to wait and see how successfully the team can now execute. With IBM, IonQ, and others striving for comparable achievements — but using better-established methods for generating their qubits — we're in something of an arms race to move past the NISQ period right now.

Microsoft also announced Azure Quantum Elements, a platform for accelerating scientific discovery by combining high-performance computing, AI, and quantum, as well as Copilot for Azure Quantum, a specially trained AI model that can assist scientists (and students) in generating quantum-related calculations and simulations.