Intel Sets Out Quantum Ambitions

Tapping into its decades of experience in semiconductor technology. Intel launched its 12-qubit silicon quantum chip, Tunnel Falls, in June 2023, making it available to the research community.

In this Q&A, Intel’s director of quantum hardware, Jim Clarke, discusses how his background in lasers set him up to lead the division, Intel’s full-stack quantum ambitions and why, despite starting with fewer qubits than some competitors, Intel is optimistic about scaling.

Enter Quantum: How did you start working in quantum computing with Intel?

Jim Clarke: I've been at Intel for 23 years and my background was studying reactions using lasers, what we would call spectroscopy, and the reactions have a very quantum mechanical component.

When I came to Intel I first worked in lithography, or using lasers to pattern wafers, and then moved to interconnect research, the wiring between devices. In late 2015 I launched a program on quantum computing here and since then, I've been managing that.

Our team consists of material scientists on how to make the best devices, device engineers, measurement engineers and quantum physicists, and we have a theory team as well. On the hardware side, we're well-positioned to build the best devices.

There are many different types of qubit devices being studied throughout the world. You may be familiar with superconducting qubits, trapped ion qubits, atoms or photonic qubits. All of these are very interesting qubits, but only one looks like a transistor.

Related:Intel Officially Launches Quantum Software Development Kit

We’ve been working on transistors since 1947. It's a half trillion or more dollar industry worldwide. And so what we're doing is we're piggybacking on all of the work that the semiconductor industry is doing and applying it to our quantum approach, and we feel that this is the way that we will accelerate our achievements in quantum computing.

How has Intel and the research community been using Tunnel Falls?

First of all, we're studying it ourselves for internal purposes. We’re beginning to fabricate our next chipset with a larger array, and designing our next larger array after that, so we're taking a model that's very similar to how we do advanced transistors.

Second of all, every time we make a 300-millimeter wafer, we get hundreds if not thousands of these chips. In the U.S. and around the world, professors receive funding through various programs to study quantum research and we provide our chips to them to do the work.

We’re working with about a dozen collaborators around the world and they can use the chip to study whatever aspect of quantum computing they’re interested in. They provide us with valuable feedback on our device.

What use cases are your collaborators working on?

We have two linked programs at Intel. I focus on the hardware, making the chips, and a colleague of mine focuses on putting the software on top. For the Tunnel Falls chip, we're giving it to quantum physicists to study the quantum phenomena – the actual qubits. They tend not to be studying applications of quantum computing.

We’ve also released a software developer kit that can emulate a quantum computer for various applications. We're working on linking the two and developing a full silicon quantum stack that you could then use to study similar chips like some of our competitors do.

What are you hearing from the industry about what they want from Intel in the quantum space?

As an innovator in semiconductor electronics, the community expects Intel to be a player in quantum hardware. We’re in a unique position in this regard, in that we're leaders in the hardware space and we have software expertise as well, so we can combine them into a full stack.

We have fewer qubits today, but our ability to scale will be faster. Once the hardware exists, quantum becomes a software game. It'll be an important component within the high-performance computing or supercomputing space. It will augment, not replace, a supercomputer. As Intel is already working in the HPC space, quantum computing will be an important component of that.