In my previous article on IBM Quantum computers, I wrote about IBM’s plans to improve access to its quantum computers. This article describes the update to IBM’s Quantum computing roadmap as revealed by Darío Gil, Senior Vice President, Director of Research at IBM Think in June.
IBM is building accessible, scalable quantum computing by focusing on three pillars:
· Increasing qubit counts
· Developing advanced quantum software that can abstract away infrastructure complexity and orchestrate quantum programs
· Growing an ecosystem of quantum-ready enterprises, organizations, and communities
IBM originally announced its quantum development roadmap in 2020. To date, the company has hit its planned releases on the original timeline. In addition to new quantum systems, IBM has sped execution performance by 120x using Qiskit Runtime, IBM’s containerized quantum computing service and programming model, from previous experiments.
The next step in IBM’s goals to build a frictionless development experience will be the release of Qiskit Runtime in 2022, which will allow developers to build workflows in the cloud, offering greater flexibility. Bringing a serverless approach to quantum computing will also provide the flexibility to distribute workloads intelligently and efficiently across quantum and classical systems.
To help speed the work of developers, IBM launched Qiskit Runtime primitives earlier this year. The primitives implement common quantum hardware queries used by algorithms to simplify quantum programming. In 2023, IBM plans to expand these primitives, as well as the capability to run on the next generation of parallelized quantum processors.
Quantum Hardware Scaling
Later this year, IBM is scheduled to deliver the 433-qubit Osprey quantum computer and dynamic circuits. IBM used 3D packaging to place a complex tangle of microwave circuit components and wiring on multiple physical levels close to the quantum processors, enabling the faster execution of dynamic quantum circuits. IBM’s experience packaging qubits will then enable construction of the 1121-qubit Condor computer, with minimal impact to individual qubit performance, in 2023. IBM expects Condor to be the first quantum computer with more than 1,000 qubits. After Condor, IBM will use chip-to-chip couplers to build even larger quantum systems.
“Our new quantum roadmap shows how we intend to achieve the scale, quality, and speed of computing necessary to unlock the promise of quantum technology,” said Jay Gambetta, VP of Quantum Computing and IBM Fellow. “By combining modular quantum processors with classical infrastructure, orchestrated by Qiskit Runtime, we are building a platform that will let users easily build quantum calculations into their workflows and so tackle the essential challenges of our time.”
To build this new quantum roadmap, IBM is targeting three scalability “regimes” or steps to scale its quantum processors.
The first step requires building capabilities to “classically” communicate and parallelize operations in a non-quantum way across multiple processors. This step opens the door to a broader set of techniques such as improved error mitigation techniques and intelligent workload orchestration, which combine classical compute capabilities with quantum processors.
The next step is building short-range, chip-level couplers between quantum chips. Using these couplers, multiple chips can be connected to effectively form a single larger processor. This multichip modularity is key to scaling.
Ultimately, the third step to reach larger scalability is developing quantum communication links between quantum processors. These quantum communication links connect clusters of quantum processors together into a larger quantum system.
IBM plans to be using all three of these scalability techniques by 2025 to build a 4,000+ qubit processor based on multiple clusters of modularly scaled processors.
Future quantum computing systems will be called IBM Quantum System Two. A central approach to building IBM Quantum System Two will be modularity, which will be necessary to increase the scale of IBM quantum chips in the future.
System Two introduces a new generation of scalable qubit control electronics together with higher-density cryogenic components and cabling. The platform brings the possibility of providing a larger shared cryogenic workspace, opening the door to potential linking of quantum processors through novel interconnects. System Two is a major step toward a true quantum data center. A prototype of this system is targeted to be up and running in 2023.
While building systems with more qubits is important for extending the capabilities of quantum computing, the quality of these qubits is also essential to building practical quantum computers. Qubit quality refers to the amount of time that the qubits are entangled and the error rate of the results. IBM has a metric for qubits called Quantum Volume (QV). IBM says its quantum systems are moving from a QV of 256 last year, to a QV of 1024 this year. The Falcon r10 system has under a 1 in 1000 error rate today. IBM handles its error management in the Quiskit Runtime.
There is progress being made with error mitigation and suppression techniques to improve the ability of quantum software to minimize the effect of noise on the users’ application. These are important steps on the path towards the error-corrected quantum systems of the future.
The next step to scaling quantum computers will be to make quantum communications links between chips and between cryostats. First, IBM plans to connect three or more Heron 133 qubit chips using classical (non-quantum) logic connections in 2023. With classical interconnects, the quantum state must be resolved to a binary logical result. But with the Crossbill quantum computer in 2024, IBM plans to interconnect chips with quantum entangled connections, which communicate in a quantum state. The connection between three chips should deliver 408 qubits. IBM will offer both system scaling options for experiments.
In addition to the potential for using quantum computing to solve complex problems, this technology can also be used to crack today’s data encryption. While some cryptographers are skeptical that quantum computing can reliably be used to break cryptography within the next decade, IBM is already planning to mitigate the issue by offering quantum-safe cryptography. For example, the recently announced Telum Z16 mainframe has quantum-safe encryption.
IBM continues to leverage its traditional computing, quantum expertise, packaging technology, extensive software resources, and new business models to expand the developer reach and market opportunities for quantum computers. IBM’s super-cold qubits are also fast – 1,000 times faster than Ion-trap quantum computers. The company has committed to scaling quantum computing and adding greater capabilities over a multi-year roadmap.
More information about IBM’s Quantum Research is found at:
Tirias Research tracks and consults for companies throughout the electronics ecosystem from semiconductors to systems and sensors to the cloud. Members of the Tirias Research team have consulted for IBM, Nvidia, Qualcomm, and other companies throughout the AI and Quantum ecosystems.