Why Industry Should Pursue Quantum Results, Not Quantum Supremacy

QuEra makes quantum computers that use qubits based on neutral atoms. The company recently celebrated the first anniversary of making its technology accessible via the AWS quantum cloud computing service Amazon Braket. It is the only known publicly accessible neutral atom quantum computer in the world.

In this Q&A, QuEra chief marketing officer Yuval Boger discusses the benefits and use cases for neutral atom quantum computing, and why achieving quantum results is a better goal than pursuing quantum supremacy.

What are the advantages of neutral atom quantum technology?

Neutral atom computers work at room temperature; you don't need the big chandelier or cryogenic cooler. When you go deeper into the machine, neutral atoms are our qubits. The atoms are held in place by tiny laser beams that can move them. This achieves all-to-all connectivity, which means that every qubit can interact with every other qubit. In a superconducting system, every qubit can only interact with two or three neighboring qubits.

When we talk about scaling up to a larger number of qubits,  superconducting quantum computers require optical fiber interconnects between modules and that makes it really difficult. That's not the case with neutral atoms. We and others have shown that even 10,000 qubits would not need these interconnects.

Related:Quantum for Bio Team Targets Treatment for Muscle-Wasting Disease

And lastly when you think about scaling, for superconducting qubits every qubit needs two or three control wires. So if you have 10,000 qubits, you're going to need 30,000 control wires. For neutral atoms, because they can move around, you need dozens of control wires for the entire system.

What problems are your customers solving with your technology?

We see three types of problems that are being solved. The first one is simulation, which is primarily used by those who are interested in the quantum-level microscopic behavior of materials.

The second is optimization. BMW is one of our clients, and they published several papers on using our computer for hybrid optimization where part of the optimization is done on our quantum computer and the other part is just done on a classical computer. J.P. Morgan has also run optimization problems on our machine.

The last one is machine learning. Deloitte Consulting used our machine-learning pipeline to analyze medical images and images from a smart factory that they own, and they got results that were better than what they were able to achieve classically.

We’ve also worked with energy companies and financial services companies and received grants from DARPA to deal with molecular analysis and the prediction of severe weather events.

You’re on record as saying we should be pursuing quantum results instead of quantum supremacy. What do you mean by that?

The 2017 Google quantum experiment solved a useless problem, but in a way that no quantum no classical computer could solve. It leads to healthy competition, but often someone can develop a classical algorithm that does the same based on that work.

When you talk with corporate customers, they don't care about world records, they just care about delivering value to their company, and that value could come in two ways.

One is having a quantum computer that performs better than when we can do it classically, maybe it optimizes quicker or provides a better solution. Another interesting approach is if it delivers the same performance but it uses a fraction of the power; supercomputers and HPC centers are using too much power.

We’re starting to see quantum computers can start delivering results that are better than simple classical algorithms. We're not claiming quantum supremacy yet, but I think it's heading in the right direction.