IBM has announced that its quantum system has achieved a 120-fold increase in the speed at which it can simulate molecular activity. Previously, the company had planned to speed quantum workloads up to 100 times on its current roadmap. The series of improvements include algorithm improvements, processor performance improvements, control system improvements, and Qiskit Runtime, a quantum computer container service.

Translated from: IBM official blog

IBM quantum company will workload by | 120 times the speed of IBM research blog

Last fall, we made an ambitious commitment in IBM’s Quantum Roadmap for scaling quantum technology to demonstrate a 100-fold increase in quantum workloads. Today, we are pleased to announce that we have not only achieved that goal, but defeated it. Thanks to a series of improvements, including leveraging Qiskit Runtime’s ability to run quantum programs entirely on the cloud, the project’s results show a 120-fold increase in simulated molecular speed.

So far, we have focused primarily on performing quantum circuits or sequences of quantum operations on IBM quantum systems. However, practical applications also require a lot of classical treatment. We use the term quantum program to describe this mix of quantum circuits and classical processing. Some quantum programs have thousands or even millions of interactions between quantum and classical. Therefore, it is crucial to build systems that can speed up the execution of quantum programs, not just quantum circuits. Systems built to execute quantum programs need to have significantly greater effective capacity, and they need to be improved throughout the stack, including cloud service design, system software, control hardware, and even quantum hardware.

Back in 2017, the IBM Quantum team demonstrated that quantum computers can simulate. Note 1: Simulations of beryllium hydride are also detailed in this Nature cover story, “Hardware Efficient Mutant Quantum Solver for Small Molecules and Quantum Magnets.” Lithium hydride molecule 1– this is a preview of the kinds of applications we hope quantum computers will solve in the future. However, the process of modeling lithium hydride molecules with today’s quantum computing services takes 45 days because the circuits are passed back and forth between classical and quantum processors, introducing significant delays.

Now we can solve the same problem in just nine hours — 120 times faster.

Figure 1: Qiskit Runtime workflow.

There are a number of improvements involved in this feat.

  • The algorithmic improvements have reduced the number of iterations of the algorithm needed to reach the final answer by a factor of 2 to 10. System software improvements reduced each iteration by about 17 seconds.

  • Improvements in processor performance resulted in a 10-fold reduction in the number of takes or repeated circuit runs required for each iteration of the algorithm.

  • Finally, improvements in control systems, such as better readout and qubit reset performance, reduced the time per job execution (i.e., dozens of circuits per batch) from 1,000 microseconds to 70 microseconds.

  • The final push came with the introduction of Qiskit Runtime, a containerized service for quantum computers. As code is passed between a user’s device and a cloud-based quantum computer, developers can run their programs in the Qiskit Runtime execution environment instead of handling the work for them in the IBM hybrid cloud. The new software architecture and OpenShift operator enable us to maximize computation time and minimize wait times.

We hope this acceleration will allow more developers to experiment with quantum applications in chemistry — and beyond. For example, Qiskit Runtime will allow users to try out our powerful new quantum kernel alignment algorithm, which searches for an optimal quantum kernel to use for machine learning tasks. We have recently used this algorithm to show that quantum computers will exhibit more rigorous speeds in supervising machine learning than classical computers.

The IBM Quantum team is committed to finding practical quantum computing use cases and making them available to as many developers as possible. We hope the Qiskit Runtime will enable users around the world to take full advantage of the 127-qubit IBM Quantum Eagle device due out this year — or the 1121-qubit Condor device planned for 2023.

Qiskit Runtime is currently being tested for some members of IBM’s quantum network. Fill out this form to learn more!

IBM quantum. Today’s tomorrow counts

Quantum starts here