QCI CEO on Optimization Using Quantum Computers

Quantum Computing Inc. (QCI) announced a partnership with quantum photonic technology specialist in February, merging QCI’s quantum software solution, Qatalyst with QPhoton’s hardware to jointly develop business solutions.

QCI caused a stir in August when it used its first quantum computer to solve a “3,854 qubit” complex optimization problem for BMW on the placement of vehicle sensors for autonomous driving in just six minutes.

More recently, the company forged a partnership with the Virginia Innovation Partnership Corporation (VIPC) to determine the best flight trajectories for drones using QCI’s Qatalyst software and Quantum Photonic Systems hardware and launched a subscription service to its Dirac1 quantum computer.

Last week the company announced plans to construct and operate a quantum nanophotonics technology manufacturing and research center to expand its current optical chip development capabilities, aiming to tap into regional funding incentives courtesy of the CHIPS and Science Act of 2022.

QCI CEO Bob Liscouski is recognized as a security industry leader and was appointed by President George W Bush as the first assistant secretary for infrastructure protection at the Department of Homeland Security. Rounding off a busy few months for the company, in this Q&A Liscouski discusses QCI’s recent developments in detail, explains how photonic quantum computing addresses optimization problems and looks ahead to QCI’s plans for next year.

Related:CHIPS Act Helps Fund US Quantum Chipmaking

Enter Quantum: QCI recently released its roadmap which included a subscription service to access its Dirac 1 Entropy Quantum Computer. What customers are you targeting?

Bob Liscouski: Right now, they're optimization use cases. Dirac 1 is the same machine we used on the BMW sensor optimization problem, but we've been testing it and using it for different clients for different types of optimization problems.

We have a professional services group that will help a client find and orient a problem that can be presented to the Entropy Quantum Computer (EQC). Based on our internal expertise, we can do supply chain logistics and even some financial problems, which are big areas.

But with others, we don't have to necessarily have the expertise. If they have the in-house expertise to formulate the problem, we can work with them on it and then present that to the EQC for running. Once they get trained up, they can just do it on their own.

We're hoping to not limit it to a specific domain. I just want people banging out on the machine and trying to use it. We're opening it up to we the academic side eventually through our Qatalyst platform, on Brake. The other piece of this is workforce development. We plan on using that machine via subscription service bill to apply our approach on the workforce development side.

The BMW sensor placement challenge was described as a “3,854-qubit” problem. No current quantum computer has that number of qubits, so how was Dirac 1 able to solve it?

We refer to it with the same vernacular as a gate model approach but use a photonic approach. The qubits on the gate model machine and the qudits (photonic qubits) on the photonic machine are not one-to-one equivalent in terms of definition.

Photonic qudits don't need error correction, so the ability for us to be able to run these problems would be the equivalent of a gate model machine that's using 3,854 error-corrected qubits.

Automotive has seen a lot of use cases for quantum computing. Have you worked with BMW or any other automotive companies since completing the sensor placement challenge?

We've been working with some other automotive companies as a result of that, but BMW hasn’t followed up. We do believe that the sensor optimization problem is a problem that is of deep interest to the community, so I’d like to see a little bit more uptake.

When BMW described the problem, to us, it was a hard problem to solve. It was not necessarily something that they were going to look to apply commercially, they just wanted to see if a quantum computer company could solve the problem. It effectively became an academic exercise.

But the good news is, the demonstrable results we got from doing the BMW problem have resulted in clients coming to us to do other optimization-type problems, such as route optimization. The VIPC contract is a route optimization problem, looking at the capability to route drones from the depot to their delivery point.

What other quantum capabilities is QCI working on?

We have a truly random quantum number generator that we’re planning on moving into the marketplace that I'm going to be talking about that at a cybersecurity conference at the end of this month. We also can provide private peer-to-peer authentication and a network on top of that. We're attacking the cybersecurity components from the various layers; it's not all about encryption, it's about the transport and authentication layers.

The quantum random number generator is really interesting because it has much more than just a cybersecurity application, it can also do simulations for financial insurance applications.

We're also engaging in a negotiation for a government contract to provide our LIDAR capability to be used to measure various terrestrial features from space.

For a small company, this combination is the secret sauce. It wasn't just the acquisition of QPhoton, it was a combination of our software capabilities and the QPhoton capabilities that combine to allow us to be able to take these things into the marketplace very quickly.

What are QCI’s plans for the year ahead?

We're continually improving the Dirac 1 and we expect to be putting out into the field a Dirac 2, which will be a new improved version with augmented optimization capabilities. And I think one of the later models that we continue to work on will be able to do more AI and ML work and some of the other areas of interest.

As a result of the CHIPS Act, we've been engaging with several states around the United States that are interested in doing quantum chip development and sponsoring us to develop a lab. I've been actively negotiating and discussing these opportunities with the states. The CHIPS Act is intended to finance a lot of the reshoring of companies back to the US to develop chips, but there's also an R&D component to that sets aside a significant amount of investment for technologies exactly like ours.

We do the chip development for our quantum units, and that’s currently onesie-twosie. A quantum chip is different from a silicon chip, it requires a different process layer on top of a more traditional manufacturing process. That's a big deal for us and we expect to have our own facility to develop chips for our own use and ultimately commercialization.