QUANTUM COMPUTING is one of those technologies that seems forever just over the horizon, much like the flying car or an iPhone that lasts all day on a single charge. Developments come and go, announcements are made, each sounding crazier than the last, but tangible benefits never seem to appear.

Physicist Richard Feynman, one of the first to conceive of a quantum computer, due to his interest in the relationship between physics and computation, often has a quote attributed to him—it’s possibly apocryphal, but most certainly apt: “If you think you understand quantum mechanics, then you don’t.” The same is almost certainly true of quantum computing, where there are so many new terms and variables to consider, it seems impossible for a single human brain to hold them all. The fact that so many of them sound like science fiction is perhaps part of the discipline’s appeal.

Having multiple approaches to solving the same problem certainly keeps the scene lively, and we envisage endless debates between scientists along the lines of whether Intel or AMD is best, and whether the cryogenic cooling plant really needs all that RGB. It also means that we’ve had a solid run of news announcements and tech demonstrations recently, as work held up by the pandemic begins to bear fruit.

BENDING THERMODYNAMICS PROBABLY...

Take time crystals. Yes, really. Despite sounding like something the Emperor Zog is searching for, protected by a plucky band of heroes including at least four kids and a puppet, a time crystal is a real thing: Google has made them using a quantum computer.

Emperor Zog can relax, however. A time crystal has possible practical uses as quantum memory or as a sensitive detector of quantum fields but isn’t going to destroy the universe. Imagine a system of particles in its lowest energy state, meaning it cannot lose any more energy to its environment, yet remaining in motion. The particles cannot come to rest because they are already in their quantum ground state, and usually would be still, but remain moving. It sounds like perpetual motion does an endrun around the second law of thermodynamics, but it has been demonstrated by a team from Stanford, MIT, and Google using Google’s Sycamore quantum processor.

Being in motion means it can flip between two states without losing energy, in theory doing this forever like a pendulum that will never stop swinging. MIT physicist Frank Wilczek wasn’t involved in the work, but he hypothesized the existence of time crystals back in 2012. “They can be sensitive probes of certain kinds of external fields, so they will give us, in principle, new kinds of exquisitely sensitive devices,” he told New Scientist.

The other thing about quantum computing is that there are so many approaches. It’s as if the early pioneers of our CPUs couldn’t agree on a material to make transistors out of, and instead split into factions, one using silicon, another using luminiferous ether, and another using cheese. They all work, and arguments persist about which is the best way of creating one.

Google’s Sycamore computer is an example of one kind of processor, the superconducting quantum processor. This is the sort of technology most commonly associated with quantum computers and requires cooling systems that chill them to within a fraction of absolute zero. Under these conditions, the qubits— the quantum equivalent of bits in a classical CPU—become superconductors that allow electrons to flow freely, scaling up the curious behavior of quantum mechanics to enable the computer to work. Microwave pulses are used to vibrate the qubits, and when two neighboring qubits reach the same frequency, they become entangled. This means that measuring the state of one tells you about the state of the other. Einstein famously poked fun at this idea, dismissing it as “spooky action at a distance,” but since his death, it has been proved to be a real phenomenon.

QUANTUM SUPREMACY

There’s another reason for keeping your computer cold, it blocks out the effects of the outside world. Noise is the great enemy of quantum computing, and noise-free, error-corrected qubits are its holy grail. That’s not to say noisy qubits aren’t useful for some things, Google was able to claim quantum supremacy (where a quantum computer solves a problem no classical computer could solve in a feasible amount of time) with its Sycamore chip in October 2019, even though its 54 qubits (it was meant to be 55, but one broke) aren’t fully error-corrected.

Google claimed that its computer could perform a random sampling calculation, where it verifies that a list of numbers has been randomly generated, in three minutes and 20 seconds. This task would have taken IBM’s Summit, the most powerful classical supercomputer, over 10,000 years.