- IBM’s roadmap for quantum computing leans heavily into software/development as the interface to make solving problems on quantum computers as easy as it is on classical systems.
- IBM’s introduction of its modular quantum computing plan will allow, in the long term, parallelization of quantum computing, essentially creating larger quantum CPUs out of several smaller ones.
As it has been for the last several years, the quantum computing market is again boiling with a great deal of vim and vigor. In May 2022, IBM made an announcement that it is expanding its roadmap for quantum computing, particularly at a large scale with a focus on realistic problems. IBM’s very near-term, previously disclosed roadmap includes the 433-qubit processor named IBM Osprey, which the company expects to make available later this year. In 2023, IBM intends to introduce IBM Condor, a quantum CPU that reaches 1,000+ cubits.
IBM also announced roadmap improvements to its Quiskit Runtime, which is the company’s containerized quantum computing service and programming model. In 2023, it will add workflows to the cloud and move the Quiskit Runtime to a ‘serverless’ model; this will allow easier distribution of workloads across quantum and classical computing systems. IBM also plans to expand the primitives that encapsulate quantum hardware queries to include the ability to run them on parallelized quantum processors, which will allow IBM to deliver Quantum Serverless into its core stack in 2023. This is all part of efforts to create an environment that can switch between classical and quantum computing in a way that emphasizes the advantages of each system respectively.
The big part of IBM’s announcement, mostly due within the 2025 timeframe, is around what it calls ‘modular quantum computing,’ which consists of three parts: The first involves the ability to parallelize work and communicate across processors, both quantum and classical; this will allow more techniques for workload orchestration, and most importantly, error mitigation. The second, and one of the more exciting roadmap items, involves short-range chip-level couplers that can connect multiple chips to form a single, larger processor; IBM rightly contends that this kind of modularity is key to creating a solution at scale. The third part is around creating quantum communication links between quantum processors to form a much larger quantum computing system; IBM’s stated 2025 goal is a 4,000+ cubit processor, created modularly in clusters.
The final part of IBM’s announcement is IBM Quantum Safe, which is a new suite of cryptographic technologies and consulting services that help companies address the security issues of current cryptographic techniques in the coming quantum era. Firstly, IBM plans to offer education with regular updates on insights to migrate to quantum-safe cryptography. Secondly, it will provide strategic guidance via its IBM Quantum Safe Scope Garage workshop, offering starting guidance and education for quantum-safe encryption in the enterprise. Thirdly, IBM will offer risk assessments and discovery so a company can know its current cryptographic stance and risks to quantum cryptography attacks when those become viable. Lastly, IBM will enable migration to agile and quantum-safe cryptography. Together, these are designed to help customers and take a realistic approach to what real quantum supremacy means to enterprise and government encryption practices and techniques.
IBM’s approach has been, and continues to be, practical and multi-pronged. It has – just as many quantum computing companies have been doing – put emphasis on the software stack as a vital piece that will make wide-scale adoption possible. Of course, it has also been pushing the quantum computing hardware itself with its large engineering resources. IBM is sensibly addressing customer fears and helping them plan for the impact of quantum supremacy on the cryptographic systems widely in use today. IBM brings a great deal of practicality to the Wild West that is quantum computing, and the approach should be emulated by other quantum computing companies looking for long-term success.