IBM recently announced Version 1.0 of its Qiskit quantum computing software, but it does not represent the computing giant’s first stab at quantum hardware or software, as the company has been developing quantum processors and full-scale quantum computers, and the programming tools and software to go along with them, for many years.
Rather, Qiskit 1.0 is both a collection of everything IBM has been doing in quantum computing software for the last seven years since it debuted its earliest Qiskit software development kit tools, as well as an expansion of these components into a full quantum computing software stack. That stack includes services that reflect the roles that classical computers will continue to have in a quantum world, and that still-new generative AI technology will have in helping to program quantum computers.
Jerry Chow, IBM Fellow and Director of Quantum Systems and Runtime Technology at IBM, said during a launch briefing for Qiskit 1.0 that the platform has had more than 600,000 users since IBM debuted it in 2017 to help early quantum computing developers and researchers work with IBM’s then-fledgling 5-qubit processor, which it had allowed cloud access to right around the same time. Over the last seven years, those users “have run over three trillion circuits on IBM Quantum computers using Qiskit. Researchers have published nearly 2,900 research papers with Qiskit and IBM systems, and over 700 universities around the world have developed quantum computing classes using Qiskit. It has become the preferred open source software development kit for 69% of respondents to the Unitary Fund’s 2023 Open Source Software Survey,” Chow added in a related blog post.
The capabilities collected in this version include some basics (relatively speaking), like support for building and optimizing quantum circuits, but Chow and other IBM officials called out some other key capabilities, including:
-
AI-powered optimization of quantum circuits for quantum hardware, embedded in the Qiskit Transpiler Service.
-
Simplified execution modes for the Qiskit Runtime Service which can be tailored for performant execution of quantum circuits on quantum hardware.
-
The Qiskit Code Assistant, powered by watsonx-based generative AI models, to
automate the development of quantum code.
-
The Qiskit Serverless open-source tool to run quantum-centric supercomputing
workloads across quantum hardware and classical clusters.
Blake Johnson, IBM Distinguished Research Scientist and Quantum Engine Lead, said capabilities that bring AI-driven programming to the table already are making a difference. “The transpiler service with AI instruction approaches is giving us about a 20% to 25% benefit… in terms of producing circuits with fewer total instructions as well as less, doing them in less time, allowing us to execute jobs faster,” he said.
Chow added that IBM has been experimenting with AI-enhanced transpilation methods based upon a reinforcement learning approach. “Already this exploration is yielding promising results, with a reduction in both the number of two-qubit gates and circuit depth on a variety of benchmark circuits,” he said. “We are making these advanced transpilation capabilities available to our Premium users with the beta release of the Qiskit Transpiler Service.”
In addition, the Qiskit Serverless capabilities come at a key time, as quantum simulations continue to be run on classical computers, and quantum computers are in some cases being deployed alongside classical supercomputers for mutual benefit.
As IBM’s main quantum computing software tool kit, Qiskit needs to do a variety of things well. “It needs to leverage advanced classical and quantum computation,” Chow said. “It also needs to be able to optimize a problem for efficient execution on a quantum system, and then actually execute the quantum circuits on the actual real systems. We've been developing a lot of these tools over the past few years and now it’s really coming together to comprise this much larger Qiskit software stack.”