Matthias Steffen, IBM Physicist: “You can almost feel that we’re approaching quantum advantage.”

Matthias Steffen completed his physics degree at Emory University (Atlanta) 27 years ago, earned his PhD in electronic engineering at Stanford, and has dedicated his life to the discipline meant to explain matter and revolutionize the technological world. He joined IBM in 2006 and has led the Quantum Processor Technologies organization for the past 15 years. He is responsible for the design, construction, and testing of advanced large-scale systems to bring about quantum advantage and fault-tolerant computing. Steffen has announced the Quantum Starling, a device expected to revolutionize this field as the world’s first large-scale quantum supercomputer.

Question. What does Quantum Starling signify?

Answer. In this field, we all face how to deal with the errors that quantum computers are highly prone to. We need to find error correction codes to exponentially reduce error rates with a slight increase in resources. What our team has discovered is the LDPC [Low-Density Parity-Check]. With this and new technologies in the hardware, we have a plan that allows us to build Starling, which will have 200 logical qubits and we will be able to perform 100 million quantum operations by 2029.

Q. Can we already talk about quantum advantage, resolving problems significantly faster or more efficiently than any known classical supercomputer?

A. It will mark the beginning of large-scale problem solving. We are quite confident that we will be able to tackle interesting problems that offer quantum advantage. We are already able to resolve the ground state of some complex molecules and compete with the best-known classical methods.

Q. Will a single unit of Starling be 20,000 times more powerful than any existing quantum computers?

A. The error correction codes allow for exponential error suppression using small groups of physical qubits and will enable us to execute 100 million gates. That is really the true power. Reducing errors to something incredibly small allows us to execute 100 million operations with 200 logical qubits.

Q. Starling will be operational in 2029. Is any prototype already being tested?

A. We are building small prototype structures. This particular error correction code requires additional hardware development that did not exist until the beginning of this year. We have multiple test devices where we have coupled one qubit to six others without observing any reduction in performance compared to our current processors. High-quality long-distance connections are also required, without performance degradation. Additionally, we have improved the design methodology. The test structures we have built in the lab incorporate all components tested individually. IBM has a long tradition, not only in design but also in semiconductor manufacturing, and this is great.

Q. Will Starling be accessible via the network?

A. Our goal is to make all these devices and systems available to customers.

Q. Does it combine artificial intelligence or is it just quantum technology?

A. Quantum computing is a method for solving certain problems. It is not a magic wand for solving all the world’s problems. We know there is an exponential advantage in the correct use of quantum computers, and this is the really exciting part that fits very well with the rest of our business model. IBM is a hybrid cloud and artificial intelligence company, and the intention here is to provide computing solutions to customers in the form of hybrid cloud and AI and quantum solutions. Our intention is to seamlessly integrate all computing technologies.

Q. What would you like to see discovered thanks to Starling?

A. I’ve been in this field for 27 years, and while quantum computing has existed for quite some time, it has only been in the last decade that we have transitioned from physics experiments to engineering. It has been an absolutely incredible journey. Some recent experiments are nearing quantum advantage. You can almost feel that we’re getting there. What I would love to see on this journey of making devices a reality is some chemistry-related application, which is something we’ve already been doing, so we can find new compounds for disease treatment.

Q. Does Starling advance IBM’s roadmap toward effective and fault-tolerant quantum computing by almost half a decade?

A. The discovery of LDPC, which requires far fewer qubits, really helps reduce engineering overhead compared to having to build something with 20 million qubits. Producing this type of structure is incredibly difficult. It was necessary to reduce the number of qubits from a performance perspective and make it modular. Another thing that has accelerated the path is that we can now go from design to testing a device, even with very complicated structures, in a matter of months. Additionally, IBM has decades of experience in semiconductor manufacturing.