IBM announced its path to achieve over 100,000 qubits and over a billion circuit gates. When realized, IBM may create the world’s first platform for universal computation in a quantum system. It sounds like Quantum Nirvana is finally in sight.
Building a practical quantum computer remains elusive to some of the most advanced research teams in the world. Unlike classical computers that rely on bits, which can represent either 0 or 1, quantum computers utilize qubits that can exist in a superposition of both 0 and 1 simultaneously. This unique property allows quantum computers to perform computations exponentially faster than classical computers for specific problems. The challenge now is to increase scale and reduce errors, which can be
However, harnessing the power of quantum mechanics is no easy feat. Quantum systems are extremely cold (the average temperature in deep space is 2.7 Kelvin while quantum processors must be at 15 mK ), inherently fragile and susceptible to decoherence, where interactions with the environment disrupt the delicate superposition states of qubits. Additionally, manipulating and measuring qubits with precision is a significant challenge, as any disturbance can introduce errors that accumulate and compromise the integrity of the calculation. In order to scale, these error rates must drop from one in a thousand to one in a million errors per calculation.
Quantum System Two
Despite these formidable obstacles, researchers worldwide are progressing steadily in overcoming these challenges, with IBM Research apparently in the lead. At the IBM Quantum Summit, researchers announced a brand new Quantum System Two, built on three IBM Heron Processors, and for the first time, shared a roadmap that will deliver quantum at scale with error mitigation and correction and the software needed to make these systems useful.
IBM says we are currently in Era 2 of 3. Research in this era focuses on characterizing quantum hardware, error mitigation and suppression, and proof-of-concept applications. Having published some 2595 research papers that share their ideas and progress, IBM will now set up eight quantum computing centers in the US, Canada, Japan, and Germany by the end of 2024 to provide access to System Two for scientists and partners worldwide. Era 3 will expand the scale and deliver error correction.
The Future of Quantum at IBM is nearly at hand.
While Quantum System 2 is a significant leap, the real excitement occurs when that technology can scale while massively reducing the error rates. IBM said they think error correction is closer than they had imagined. Much of this confidence comes from new research beyond error mitigation we reported a few months ago. When coupled, literally, with new interconnect technology to enable scaling, IBM can now see the goal line.
IBM’s new Quantum roadmap extends beyond the hardware. It details the software and enabling hardware technology needed to deliver quantum advantage, when a quantum system can solve problems that traditional one-and-zero computers simply cannot solve in any amount of time. Note the little green circled checkmarks that mark achieved milestones.
While IBM would not confirm that “Blue Jay” is named after Jay Gambetta ;-), the scientist who leads the company’s quantum lab, this platform will deliver the long-sought goal of quantum research, solving a vast array of problems traditional computers cannot. After all, that’s the whole reason the industry is spending billions of dollars to build this cold future. Unsolved problems in AI, chemistry, financial services, life sciences, physics, and basic research could finally become tractable, making results close to humanity’s reach.
However, along the way, IBM has provided yearly goals for each of the major components it is researching. New error correction codes and system modularity, with interconnected building blocks, will enable millions of qubits to provide the same overall performance with error correction code presenting as hundreds of thousands of qubits, interconnected with new couplers IBM is innovating.
Below is a blow-up of the key steps in the path to Blue Jay. Kookaburra is the next major chip in 2026, forming the initial building block on which scalability and error reduction will evolve into Blue Jay in 2033.
IBM has yet to miss a milestone goal in its quantum computing program, and we believe they aren’t about to start now. This remarkable track record has given computational scientists the tools to advance their work and IBM management the confidence to publicly announce a roadmap to achieve quantum nirvana.