memQ CEO argues quantum networking is the missing link to scaling quantum computing

[Gareth19]

Electronic Design opinion piece from Charles Foley, chairman and CEO of quantum networking company memQ, arguing that networking is the missing piece in scaling up quantum computers. Foley has spent more than 40 years in tech leadership, including time at IBM and NetApp. His main point is that the industry should stop waiting for single giant quantum computers and start connecting smaller ones together

He draws a comparison with supercomputing. In the 1980s and 1990s, the biggest machines were single monolithic systems, but the industry later moved to clusters of many smaller computers linked together. Foley thinks quantum computing will follow the same path. Instead of trying to build one machine with millions of qubits, you network many smaller processors

The catch is that quantum networks are much harder than classical ones. You need infrastructure that can keep delicate quantum states intact and extend them between connected systems. Improving the distance over which that works is one of the key technical challenges. He also points to US initiatives like DARPA's HARQ project and a recent executive order as signs that policymakers see networking as essential, not an afterthought

He mentions that around $4 billion was invested in quantum computing in the past year, and that commercial deployment is not expected until the next decade. The industries he highlights as likely winners are pharmaceuticals, financial services and materials science. None of that is surprising, but it is useful context

Obviously the CEO of a quantum networking company is going to argue that quantum networking is essential, so take it with a pinch of salt. That said, the logic of distributed computing is hard to argue with, and several big players are working on linking processors already. Do you think networked quantum computers are the way forward, or will monolithic machines win out?

Quantum Networking: The Missing Link to Scale Quantum Computingwww.electronicdesign.com

[EdgeRatedR]

Would love to know more about what memQ makes. The article is heavy on vision and light on product. A bit more detail would have made it more convincing

[Freddy93]

Commercial deployment next decade is a refreshingly honest timeline. Most company pieces promise far sooner

[SuperPosition24]

Distance is the real limit. Even short links between fridges in the same room are difficult right now

Long distance quantum networks need repeaters that do not fully exist yet. That is a decade of work at least

[Quarry18]

Pharmaceuticals always top these lists of likely winners. Drug discovery needs accurate simulation of molecules, which quantum machines should handle naturally

Whether networking is what unlocks that is less clear. A single good machine with error correction might be enough. Time will tell which approach arrives first

[Christopher_27]

The supercomputing comparison makes a lot of sense. Nobody builds one giant processor any more, they build thousands of smaller ones and connect them. There is no reason quantum should be different in the long run. The physics of linking qubits is the hard part. Once that is solved, scale out becomes the obvious route

[Cass64]

IBM's roadmap already includes linking multiple processors together, so this is not a fringe idea. Several of the big players are betting on modular designs. The question is whether the links can be made fast and reliable enough. Losing quantum information in transit would defeat the point

Photonic interconnects seem the most likely answer. Converting between qubit types and light is still inefficient though

[Brad]

Four billion dollars a year sounds like a lot until you compare it with AI spending. The big tech firms spend that on AI infrastructure in a few weeks. Quantum is still a niche by comparison

[Always_Reuben87]

CEO of networking firm says networking is the answer. Hard to be shocked by that one. Still a fair argument though