Quantum Data Centers Await Use Cases and Tech Maturity

Artificial intelligence-ready data centers and AI factories dominate the spotlight now, but building the digital infrastructure for quantum computing is just as critical. Yet, quantum systems, uniquely suited to solving complex problems, lack compelling business use cases to drive widespread adoption. Technical challenges also persist: quantum hardware must be miniaturized to fit standard data center racks, and significant refinements are necessary for photonic interconnects, essential for linking quantum machines or hybrid quantum-classical systems.

See Also: OnDemand | Identity Crisis: How to Combat Session Hijacking and Credential Theft with MDR

In an interview with Information Security Media Group, Lawrence Gasman, founder of Communications Industry Researchers, or CIR, explained why the quantum ecosystem is not yet ready for enterprise use today and is more suited to the scientific, research and development communities.

Gasman is a veteran analyst and author specializing in emerging technologies, such as quantum computing, advanced photonics and next-generation data centers. He is also the co-creator of the Inside Quantum Technology, or IQT, conference series. He is known for translating complex technical subjects into actionable insights for executives, investors and engineers. Gasman also writes a weekly Substack newsletter on data center technology called “From My Desk.”

Edited excerpts follow:

Can you tell us about CIR and your role there?

CIR is an industry analyst firm formed more than 40 years ago. Today, most of what we do revolves around providing forecasts and opportunity analysis on the latest technologies for data centers, such as Ultra Ethernet, co-packaged optics, or CPO, quantum computing, and the latest powering and cooling options. In these and related areas, CIR publishes reports and carries out demanding consulting assignments for companies and government departments.

With Microsoft’s Majorana 1, AWS’ Ocelot and Google’s Willow chips announced this year, do you see quantum computing becoming mainstream as hardware form factors shrink?

Form factor, if you mean the size of the machine, absolutely. I think there will be smaller QPUs (quantum processing units) aimed at data center racks and HPC (high performance computing) applications.

The Majorana chip is not necessarily an example of that. Microsoft and Google are more about building the world’s most powerful quantum computers. Microsoft has always said to do real quantum computing, you need a million qubit machines. That’s where they’re heading with their Majorana chip. The reason for that is the kind of applications you’re going to want to run. In particular, quantum chemistry needs that kind of capacity.

There’ve been some papers recently. One of them, presented at our last Scandinavian conference about six weeks ago in Sweden, says quantum computers only make sense when you’re dealing with really complex problems. For easier problems, a classical computer can do a perfectly good job.

That said, data centers will increasingly deal with more complex problems. I do see small quantum computers in two, four, six racks emerging rapidly. What I’m talking about is not going to go to the home computer market, but medium to large enterprises will see a point in them.

What kind of enterprise applications do you see for quantum computers?

There are applications for drug manufacturing companies. You can launch powerful drugs with less trial and error and therefore at lower cost. We’re also talking about material science applications, such as coatings for the automotive and aerospace industries. There’s clear value in solving problems. These are all specialist applications.

When it comes to general-purpose enterprise use, a quantum data center can help solve complex problems in decision-making, business intelligence and strategic planning, especially when projections span long time horizons.

For example, if you’re modeling geopolitical and energy factors, advancements in manufacturing technology and materials, and you’re integrating all of them into a single computer model, and you need answers over the next 50 years, quantum computing becomes relevant.

Right now, if you’re spending $10 million to $12 million on a quantum computer, you might say, “That’s nice – maybe someday.” But when the price drops to $2 million for a data center-class machine, you’ll likely say, “Yes, getting an accurate answer will save us $2 million and we simply can’t solve this any other way due to the complexity of the interactions.”

Today’s fastest Ethernet can’t handle quantum computing throughput. What will replace it in the data center?

As I see it, there’s a significant intellectual gap between what most practical data center professionals focus on and the world of quantum data centers.

Ethernet has been around for 50 years, evolving through new generations, from an early one-megabit version to today’s Ultra Ethernet, which officially reaches speeds of up to 1.6 terabits per second.

Eventually, Ultra Ethernet will reach its limits, and the future lies in integrated photonics. The most prominent current approach is co-packaged optics, which combines high speeds with relatively low thermal issues. CPO achieves this by placing electronics and optics together in a single package, shortening the physical interconnect and reducing heat.

CPO first gained attention about five years ago and was a major topic at the OFC show in California, but interest faded a year later. Now, it’s back in a big way, largely because it addresses the enormous data rate and latency demands of AI data centers.

You may not see much about CPO in press releases today, but you’ll definitely hear about AI interconnects, which are essentially the same technology.

There’s also growing interest in quantum networking, which requires interconnects to link quantum machines.

What I’m seeing is that both traditional data center technologies and the emerging quantum data center space, which doesn’t truly exist yet, are converging on interconnects as a core challenge. And in the quantum world, those interconnects are typically photonic.