'We're hacking the process of creating qubits.' How standard silicon chips could be used for quantum computing

Quantum Motion says that its latest experiment paves the way for large-scale, practical quantum computers.

quantum-motion.jpg

Quantum Motion's researchers have shown that it is possible to create a qubit on a standard silicon chip.  

Image: Quantum Motion

Forget about superconducting circuits, trapped ions, and other exotic-sounding manufacturing techniques typically associated with quantum computing: scientists have now shown that it is possible to create a qubit on a standard silicon chip, just like those found in any smartphone. 

UK-based start-up Quantum Motion has published the results of its latest experiments, which saw researchers cooling down CMOS silicon chips to a fraction of a degree above absolute zero (-273 degrees Celsius), enabling them to successfully isolate and measure the quantum state of a single electron for a whole nine seconds. 

The apparent simplicity of the method, which taps similar hardware to that found in handsets and laptops, is striking in comparison to the approaches adopted by larger players like IBM, Google or Honeywell, in their efforts to build a large-scale quantum computer. 

SEE: Building the bionic brain (free PDF) (TechRepublic)

To create and read qubits, which are the building blocks of those devices, scientists first have to retain control over the smallest, quantum particles that make up a material; but there are different ways to do that, with varying degrees of complexity.  

IBM and Google, for example, have both opted for creating superconducting qubits, which calls for an entirely new manufacturing process; while Honeywell has developed a technology that individually traps atoms, to let researchers measure the particles' states. 

These approaches require creating new quantum processors in a lab, and are limited in scale. Intel, for example, has created a 49-qubit superconducting quantum processor that is about three inches square, which the company described as already "relatively large", and likely to cause complications when it comes to producing the million-qubit chips that will be required for real-world implementations at commercial scale. 

With this in mind, Quantum Motion set off to find out whether a better solution could be found in proven, existing technologies. "We need millions of qubits, and there are very few technologies that will make millions of anything – but the silicon transistor is the exception," John Morton, professor of nanoelectronics at University College London (UCL) and co-founder of Quantum Motion, tells ZDNet.  

"So rather than scaling up a new approach, we looked at whether we could piggy back off of that capability and use these tools to build something similar, but with qubits." 

As Morton explains, when a transistor is switched on, it sucks in a bunch of electrons that enable current to pass. Cooling down the chip to a low temperature, however, slows down this process, and enables researchers to watch the electrons as they enter the transistor one by one – "Like watching sheep walking into a field," says Morton. Instead of letting all of the particles in, the researchers allowed only one electron to enter; and once isolated, the particle could be used and measured as a qubit. 

"We're hacking the process of creating qubits, so the same kind of technology that makes the chip in a smartphone can be used to build quantum computers," says Morton. 

The significant advantage that silicon chips offer over alternative quantum approaches is scale. The qubit density that can be obtained with a silicon chip is effectively much higher due to the small size of electrons; according to Morton, this would let a single chip pack millions of qubits, where a superconducting quantum computer could require an entire building for the same yield. 

What's more, silicon chips are now sitting on decades-worth of tweaking and development, meaning that quantum devices could rely on established processes and fabrication plants. This would fast-track the development of quantum processors, while bringing down prices.  

In other words, rather than starting from scratch, Quantum Motion proposes taking the best of what is already out there. "Plus, every time the silicon industry makes an advance, you could benefit from in the qubit technology," says Morton. 

As promising as the experiment may be, it is still very early days for silicon-based quantum computing: Morton and his team, for now, have only isolated and measured the state of a single electron. In a next step, the researchers are planning on creating a quantum gate by entangling two qubits together on the chip. 

Quantum Motion's findings, rather, should be seen as a blueprint for producing quantum chips more efficiently, by leveraging existing manufacturing processes.  

The start-up's findings are likely to grab the attention of larger competitors. Intel, for one, has shown growing interest for the opportunities that silicon chips present for quantum. The Santa Clara giant has partnered with QuTech, a Netherlands-based startup, to explore the potential of silicon spin qubits.