Friday 12th August 2022

Q&A: ORNL’s Early Steps in DOE’s March to Build a Quantum Internet


Oak Ridge National Laboratory recently became one of a handful of organizations funded by the Department of Energy to develop the underlying technology required to build a Quantum Internet. The move came at roughly the same time ORNL reorganized its quantum research under a Quantum Information Science Section (QISS) headed by Nicholas Peters.

As described by Peters, “That’s sort of the biggest organization size you can have without being led by a professional manager. So I’m a scientist. My day job is running science projects. I came to Oak Ridge in 2015 and the quantum networking we had at the time was mostly focused on quantum key distribution. That work was funded by the Department of Energy’s Cybersecurity, Energy Security, and Emergency Response (CESER) office [which had] a Cybersecurity for Energy Delivery Systems (CEDS) program. That was what our quantum networking research was until a couple of years ago, when the Department of Energy’s Office of Science started funding Quantum.”

Nicholas Peters, ORNL

“The way DOE does stuff is they have workshops, you write a workshop report, and they decide if they want to make a program in it. Typically, they ask for proposals and point to a workshop report. So we did that. We started quantum networking a couple years ago and had this quantum internet workshop [to discuss] the quantum networking research that needs to still be done. So we’re a long way from having a quantum internet,” said Peters. The proposal was accepted and funding arrived last October.

It’s not hard to see why the DOE would want a working quantum internet, “It’s not a secret that DOE is investing heavily in quantum objects, developing quantum computers and things like that. If we have real quantum computers sitting at the National Labs, there’s going to be value in connecting those resources with the quantum network. We’re still a long, long, way from doing that. Right now we have the Energy Sciences Network (ESnet), which connects with a high-performance network, the national labs. If we have quantum resources sitting at the National Labs, the idea is that we’d like to have a quantum network that connects them. For example, you could take my take my special resource (photon) states at Oak Ridge and send them through the network to another national lab,” said Peters.

Presented below are a few of Peters’ comments to HPCwire during a discussion of ORNL’s quantum internet work. Much of the work is around perfecting photon sources – including perhaps “the highest bandwidth and entangled photon source ever made” – developing error correction techniques, and exploring various quantum repeater and memory technologies. The QISS section has other non-networking quantum research areas as well.

HPCwire: Maybe get us started by comparing the state of quantum computing research to quantum networking and a quantum internet.

Peters: In the U.S. in general, the government started investing in quantum computing in mid 90s. Right. And that enables us to sit here at 2022, and have a Fortune 500 company do quantum computing, right, to buy this stuff. You know Google’s got a system. Google went and bought their research group to start to start their quantum computing effort. We’re not nearly as mature in quantum internet type research right now. So you basically have to hand build everything [such as] these entangled photon sources. You can go out and buy single photon detectors as a commercial product. But you know, for a long time, you’d spend a lot of time messing with your detector to get it to work well. But now you just go get turnkey solutions.

Eventually, photon sources will do that as well. There are some companies out there selling turnkey entangled photon sources. But what we want in our photon sources is to have extraordinarily high throughput, so really bright photon sources that are really high quality. We like to have this metric called fidelity, which is this measure of state arrival. We want to try to have 99 percent fidelity, which is almost perfect in terms of the entangled state quality. We also want to have them be entangled in only one degree of freedom, because if they’re entangled in multiple degrees of freedom, then you don’t get as good a HMO (Hong–Ou–Mandel) interference. It turns out, it’s really hard to make a high quality entangled source that’s also spectrally un-entangled.

HPCwire: What’s your take on the global competition to achieve practical quantum computing and communications?

Peters: I think in general, it’s an important thing. Think about quantum [technology] in general; there are many, many countries that just can’t afford to fund a really broad portfolio of science projects. So many of them have decided, “Well, we’re going to invest in quantum because we think it’s a big deal.” We see this across the world; different countries are really focusing their science projects and programs in the quantum space. I think a lot of people think that quantum is going to be a key technology of the future. It’s going to be the kind of thing that 100 years from now is going to be all transparently underneath the way that we just live our lives and stuff will be operating and making our lives better.

HPCwire: Let’s jump into a quick overview of ORNL’s quantum networking research?

Peters: We have four projects (brief description at end of article) in quantum networking. One of those projects is an Early Career Award for one of our junior staff named Joe Lukens. We have three projects that are bigger projects, one of which is this quantum internet project. If you know much about optical networks, there’s a hierarchy and the very lowest layer is the physical layer – that’s where we’re transmitting photons on optical fibers. The quantum internet is going to look a lot like the classical internet in that you’re going to need all of the classical stuff you have out there today, but it might not be optimized the same way that it’s optimized today. What we focus on is the physical layer.

An innovative method for controlling single-photon emission for specific locations in 2D materials may offer a new path toward all-optical quantum computers and other quantum technologies. This image shows a false-color scanning electron micrograph of the array used to create place single-photon sources in epitaxial tungsten diselenide. Inset shows the Hanbury-Brown Twiss interferometry measurement proving quantum emission. Image Courtesy of Los Alamos National Laboratory.

Most of us are physicists that are working in this field, but we’ve got some computer scientists and electrical engineers working on it as well. Right now, just being able to create photons and measure them and transport them over fiber and manage that bandwidth of the entanglement is where much of our focus is. On the on the quantum internet project, our major partner is Los Alamos National Lab. They’re developing tunable sources of single photons using these carbon nanotube type technologies they’ve got. They can actually tune the wavelength of emission of these photons, which is important because typically what you’re trying to do with photons is to interfere them with each other. The type of interference you want is called Hong–Ou–Mandel interference, a really famous type of interference discovered decades ago.

HPCwire: I’ve heard of Hong–Ou–Mandel interference but don’t know much about. How is it useful in quantum communications?

Peters: [Using] Hong–Ou–Mandel interference allows you to do what is effectively a two-qubit quantum gate in linear optics. Photons naturally don’t interfere well with each other, which is good because you can send it down fiber, frequency multiplexed, and they don’t screw each other up. We have shown that you can do a complete gate-set of linear optical gates using the frequency degree of freedom [technology] that was invented at ORNL. Part of what we’re going to be doing is use that same frequency technology to do a management of the quantum signals and teleportation [i.e. quantum information sharing].

This linear optical teleportation and a two-qubit gate both use that same Hong–Ou–Mandell interference. We’re developing those [interfered photon] sources at Los Alamos. We’re also developing a special two-photon source at ORNL, which is based upon an earlier [work] that ORNL published over a decade ago, but…


Read More:Q&A: ORNL’s Early Steps in DOE’s March to Build a Quantum Internet