What quantum buyers will ask next, and how to tell who can deliver

A few years ago, almost every article about quantum computing opened the same way. Classical computers use bits that are zero or one. Quantum computers use qubits that can be both. Therefore, quantum will change chemistry, transportation, and defense.

That framing no longer works. I made this point recently in a conversation with Joshua Feinberg on the Data Center Go-to-Market Podcast (see the full episode here), and it has stayed on my mind since. The quantum industry is moving from evangelism to proof, and that shift will change what vendors say, what buyers ask, and how both sides measure success.

Why now?

Two curves are converging. On one side, the algorithms keep getting cheaper. Problems that were once thought to require enormous quantum resources now look feasible with far fewer, and AI tools are accelerating that research. On the other side, the hardware keeps getting better. The point where the two curves meet is getting closer.

That is why many timelines have shifted from "five to ten years" to "a couple of years." At QuEra, we expect our first truly useful system to be available on AWS in 2028. When a technology moves from someday to a date on the calendar, the conversation changes.

Twenty miles of range

I find electric vehicles a useful way to think about where we are.

Today's quantum computers are like an electric car with a 20-mile range. That is a real achievement, since the range used to be zero. You can drive it, get a feel for it, and learn what is different about it. But you would not use it for your commute.

The next step is a 100-mile car. It can do useful things, though not everything. Later comes the 500-mile car, and the quantum equivalent of that is when the broader set of applications opens up.

The analogy has limits. An EV's usefulness comes down to one number, while a quantum computer's usefulness depends heavily on the problem. Still, the analogy captures the key point: progress is real, it comes in stages, and each stage should be judged on what it can actually do.

The questions are changing

For years, quantum vendors competed on qubit counts, error rates, and error correction schemes. Those numbers matter, but most of the outside world cares about outcomes, not qubits.

As quantum computers move into data centers and HPC centers, buyers are starting to ask the questions they ask about any other infrastructure:

• How much power does it draw?
• How much space does it need?
• What does it cost, and what does it take to maintain?
• How do my developers program it?
• How does it integrate with my classical supercomputer?
• Who are your partners?

Some of these answers are already encouraging. Our current system draws about 20 kilowatts, a small number next to the megawatts that data center planners talk about. Installing a system at a data center in Japan, next to a classical supercomputer, taught us lessons that no qubit benchmark would capture.

The industry is starting to look less like a physics competition and more like a traditional infrastructure market. I think that is a healthy sign.

Ask for proof

If a vendor without a 20-mile car promises you a 500-mile car next year, ask a simple question: what has to happen between now and then, and what have you published or demonstrated that shows you can take those steps?

Peer-reviewed research is a good place to start. A peer-reviewed result means other scientists have examined the work and agreed that the method makes sense and the conclusions follow. Independent programs that examine technical roadmaps in depth, such as DARPA's Quantum Benchmarking Initiative, add another layer of scrutiny.

There is no shortage of quantum content. There is a shortage of realistic, honest content.

What this means for vendors

Less evangelism, more proving. Promising that quantum can solve every problem today may produce a short-term win, but it sets up a long-term loss when expectations collide with reality.

It also means being specific about what a machine is not good for, and sometimes turning down projects that don't fit today's capabilities. Successful projects build the market. Disappointed customers set it back.

What this means for buyers

The wrong reason to engage with quantum is to say on your website that you are exploring it. The right reason is a belief that useful quantum computing is now a one- or two-year journey, and that you want to be ready when it arrives.

Being ready means picking a problem that would matter to your business, building internal skills, and choosing partners who can show their work.

The bill is coming due on quantum promises. That is good news for everyone who has been doing the work.