もっと詳しく

IBM’s latest quantum computing processor marks a notable step forward for its ambitions in the field. At the IBM Quantum Summit 2022, the company announced Osprey, which has the largest qubit count of any of its processors at 433 quantum bits. That’s more than triple the number of cubits that the Eagle chip, which IBM revealed last November, has. IBM said it built on top of the 127-cubit Eagle’s architecture by keeping qubits on a single plane with the help of multi-level wiring.

“Unlike classical bits which have to be in a state of either one or zero, qubits can exist in a complex mix of both, tapping into the fundamental quantum nature of matter at subatomic levels,” IBM said in a press release. “As a result, quantum computers offer the possibility of vastly increased computing power that can be used to tackle calculations of much greater complexity in fields such as artificial intelligence, and the design of new materials for drug discovery and energy research.”

A split view of IBM's Osprey quantum computing processor, showing the layers of the chip.
Connie Zhou/IBM

The company is scaling up its quantum computing efforts with the aim of building a system with 4000-plus qubits by 2025. It’s currently on target with its roadmap. Next up is a 1,121-cubit chip called Condor that IBM hopes to debut next year.

IBM debuted an elegantly designed functional quantum computer at CES 2019. Last year, it offered a sneak preview of the design for the next-gen IBM Quantum System Two (the machine that will use its quantum processors). Now, the company has revealed more details. It designed the system to be flexible and modular with the ability to house multiple architectures and processors, and claims it will include Osprey starting next year. The company claimed the System Two design “allows for an exponential step up in quantum computing scale and enabling the vision of quantum-centric supercomputing.”

Having immense computational power at one’s fingertips doesn’t mean much if you don’t have the right software to get the most out of it. Early last year, IBM said that its Qiskit program execution environment was able to blend quantum and conventional computers to carry out complex computations in hours, when they previously would have taken months. The company said it will offer new features to mitigate and suppress errors, and provide “a fast, efficient and easy-to-use programming model for quantum computers.” What’s more, these features blend into IBM’s goal of integrating quantum computers with classical computers, including supercomputers.