How to build a quantum computer: Engineers unveil blueprint for a next-gen PC

Evan Gower February 27, 2018

Quantum computers have been proven in theory – now it’s time to get building. That’s the idea behind the University of Sussex’s efforts to lay out a blueprint to create a machine.

How to build a quantum computer: Engineers unveil blueprint for a next-gen PC

By taking advantage of quantum states, these computers will be many orders of magnitude more powerful than existing binary machines when solving certain problems. However, such a general-purpose quantum computer has yet to be built. Winfried Hensinger, professor of quantum technologies at the university, hopes to change that, revealing a blueprint that could be created using existing components, offering a practical approach to an academic problem.

We spoke to Hensinger to find out why it’s time to get practical with quantum computing and when to expect the first machine.

Why is a blueprint a breakthrough?

Quantum computing has primarily been in the field of academic research, where people look at certain effects or prove a principle. The reason people have not really been doing the engineering is it’s very difficult to build a quantum computer.

Rather than ask if it’s possible or exploring other effects or doing another proof-of-principle experiment, we’re trying to ask the question “how would you go about actually constructing a device?” We want to understand all the components: things such as energy consumption, what would be the required error rate, and what would be the technical details if we were to build a device?

How does the design work?

A key part of this design is that it’s fully modular. The size depends on the processing power you want. So certain problems would need a fairly small machine, while other problems would require a large machine.

A normal computer has a single processor, but it isn’t capable of doing many things. For bigger tasks, you have a supercomputer where you have lots of processors working in parallel. But you wouldn’t have someone building one huge processor for that… at the manufacturing capabilities or facilities where Intel makes processors, the size of what machines can process is limited.

This also applies to a quantum computer. You have a quantum computing core, but you can only manufacture that so large. You’re going to be limited by the largest wafer size a machine can process. So for us, in order to build a quantum computer that is applicable for any problem, you definitely need to make it in a modular design, meaning you can choose the number of modules according to your needs.

There have been modular designs for quantum computers, but they used photonic interconnects — optical fibre technology. We invented a new way to connect modules and that is by electric fields… this is one of the ways our approach is much simpler than any previous architectures.

Could people take these blueprints and build their own computer?

The idea is to share this to advance the technology faster and bring in quantum computers sooner. We will build a prototype and a large-scale machine, but we want others to do that too.

How long will it take to build?

It’s not like we can just have a single company and construct a device in a year — this is a huge project. We’re going to need help, especially building machines that can attack this full range of problems — it may be a bigger machine, filling a house. You can imagine we need a lot of talented engineers to build it.

At the University of Sussex, we’re constructing a small prototype device that will have two modules and fill a room but not a building — so it’ll be a fairly small prototype. At the same time, we’re going to look for partners to build a large-scale machine. We anticipate maybe ten years before it’s fully constructed and debugged.