Quantum computers can now solve problems with real-world applications faster than any ordinary computer, suggesting they could be commercially viable, say researchers at quantum computing firm D-Wave – though outside observers are more cautious.

It had long been hoped that quantum computers will be able to perform some tasks that are impractical or impossible on even the best supercomputers. Google was the first to demonstrate this “quantum supremacy” in 2019, but only for a somewhat contrived benchmark test with no practical use. Indeed, earlier this month, Google …

Google and XPRIZE are launching a $5 million competition to find practical uses for quantum computers that could actually benefit society. We already know that quantum computers can perform specific tasks faster than classical computers, after Google first claimed quantum advantage for its Sycamore processor in 2019. However, these demonstration tasks are simple benchmarks with no real-world applications.

“There’s a lot of rather abstract mathematical problems where we can prove quantum computers give very, very large speed-ups,” says Ryan Babbush at Google. “But a lot of the research community has been less focused on trying to match those more abstract quantum speed-ups to specific real-world applications, and to try to figure out how quantum computers could be used.”

To that end, Google and the XPRIZE Foundation are urging researchers to come up with new quantum algorithms as part of a three-year competition. The winning algorithms could solve an existing problem, like finding a new battery electrolyte that vastly improves storage capacity, but it doesn’t need to solve the problem in practice, says Babbush. Instead, researchers just need to show how an algorithm could be applied, detailing the exact quantum computing specifications required. Alternatively, competitors could show how an existing quantum algorithm could be applied to a real-world problem not previously considered.

The prize will judge entrants’ algorithms on a range of criteria, such as how large their impact could be, whether they tackle problems similar to those outlined in the United Nations Sustainable Development Goals, and how feasibly they can be run on machines that are available now or in the near-future.

A total prize fund of $5 million will be split into a grand prize of $3 million shared between up to three winners, $1 million shared between at most five runners-up and $50,000 for each of 20 semi-finalists.

The prize could help shift the focus of quantum computing researchers from looking at technical definitions of quantum advantage, like those demonstrated by Google or IBM, to real-world uses, says Nicolás Quesada at Montreal Polytechnic in Canada. “[The prize is] hitting the nail on the head that this is a very important problem,” says Quesada. “We need to figure out what to do with a quantum computer.”

However, finding socially beneficial quantum algorithms will require a better understanding of how quantum computers work, such as how to deal with noise and errors, says Bill Fefferman at the University of Chicago. The prize doesn’t address this foundational aspect of building quantum computers, he says.

“I’m very optimistic, in principle, that we’ll find algorithms that are really useful,” Fefferman. “I’m not as optimistic that, in the next three years, we’ll be able to discover those algorithms and then also implement them on the current hardware that will exist.”

Topics:

Nearly a fifth of all difficult-to-correct errors in quantum computers are caused by powerful particles from space hitting the machines. While these cosmic ray-induced errors have long been predicted by scientists, this first precise measurement of how often they occur could help to error proof future devices.

Ordinary computers are also susceptible to cosmic ray errors, but the delicate components that power quantum computers, known as qubits, are more at risk because their fragile quantum states are easily disturbed by…

A working, scalable semiconductor has been created from graphene for the first time, potentially paving the way for a new type of computer with greater speed and efficiency than today’s silicon chips. But despite graphene’s promise, one expert says it will not replace silicon chips any time soon.

Graphene is a material made from a single layer of carbon atoms that is stronger than steel at comparable thicknesses. It is an extremely good electrical conductor and highly resistant to heat and acids. But despite its advantages, a working graphene semiconductor, which can be controlled to conduct or insulate electricity at will, has evaded scientists. Such semiconductors are key to creating the logic chips that power computers.

The problem has been the lack of a bandgap. Semiconductors have bands of higher and lower energies and a point – known as a bandgap – at which excited electrons can hop from one to the other. This effectively allows switching on and off, creating the binary system of zeroes and ones used in digital computers.

While previous research has shown that graphene can be made to act like a semiconductor on a small scale, it had never been scaled-up to sizes that would make a computer chip practical. Previous work has shown that wrinkles, domes and holes in graphene sheets can have unusual effects on electrical flow, creating the possibility that logical chips could be made by creating the right landscape of flaws. But to date, nothing has scaled up.

Now, Walter de Heer at Georgia Tech and his colleagues have created graphene with such a bandgap and even demonstrated a working transistor. Their process should be more conducive to scaling-up because it relies on techniques not dissimilar to the creation of silicon chips.

De Heer’s group used wafers of silicon carbide that were heated, forcing the silicon to evaporate before the carbon, effectively leaving a layer of graphene on top. De Heer was not available for interview at the time of writing, but said in a statement that the electrical properties of a graphene semiconductor were far better than those of silicon chips. “It’s like driving on a gravel road versus driving on a freeway,” he said.

Silicon is cheap to manufacture and has enormous global manufacturing infrastructure, but we are reaching the limits of what these chips can do. Moore’s Law states that the number of transistors in a circuit will double roughly every two years, but the rate of miniaturisation has slowed in recent years as engineers reach circuit densities beyond which electrons cannot be reliably controlled. Graphene circuits could reinvigorate progress, but there are still hurdles to overcome.

“The fact they’re using wafers is important because that’s really, truly scalable,” says David Carey at the University of Surrey, UK. “You can use all the technology that the whole semiconductor industry is totally comfortable with to scale-up this process.”

But Carey is skeptical that the development means the world will soon shift from silicon to graphene chips, both because the new research needs lots of refinement in terms of transistor size, quality and manufacturing techniques, and because silicon has such a headstart.

“Most people who work on silicon are bombarded on a daily basis by new, wonderful materials that are about to replace it and none of it’s ever happened,” he says. “If you’re a silicon person you’re quite happily sitting on top of the mountain. The idea that I’m going to replace my laptop with graphene is not quite there yet.”

Topics:

Computers could see their performance jump by one-third through an inexpensive cooling solution based on lithium bromide salt – all while enjoying 10 times longer cooling than using alternative passive cooling systems.

To remove heat generated by a computer’s central processing units, researchers designed a system that relies on water evaporation from a solution containing lithium bromide – a salt capable of absorbing water.

Wei Wu at the City University of Hong Kong in China and his colleagues placed the lithium bromide solution within a porous membrane that only allows water vapour to pass through. They set that within a hollow plate to ensure that no salty solution seeps through to the computer’s electronics, and then added a metal heat sink to efficiently transfer heat away from the adjacent computing device.

That cooling process allows the computer processors to draw more power during intensive computing without overheating. When the computer is idling and not running hot, the passive cooling system can replenish its water and effectively reset.

“The device can spontaneously and quickly recover its cooling capacity by absorbing water vapour from the air during off hours, just like a mammal rehydrating and preparing to sweat again,” says Wu.

Testing showed that the passive cooling system could keep an off-the-shelf computer processor running below 64°C (147°F) for about 400 minutes – a huge improvement over alternative passive cooling systems. For example, one alternative uses a chromium-based metal-organic framework (MOF), which can hold a lot of water moisture. But these MOFs are expensive and only provide cooling for 42 minutes. The salt solution delivers cooling with 1000 times better cost-effectiveness than the MOF approach.

Salt cooling enabled the computer processor to improve its average input power – and therefore computing performance – by almost 33 per cent during testing.

Such passive cooling could prove extremely helpful in both personal computing devices and in computer servers packed within hot data centres. But it could also work with batteries, solar cells and buildings, says Wu.

Topics:

Artificial intelligence running on a quantum computer can now generate recognisable images of things like shoes and T-shirts, using the same methods as popular text-to-image tools like Dall-E or Midjourney. They still aren’t what you would call stunning images, but if the method scales up to more powerful machines, it should lead to much higher-resolution pictures.

Generative adversarial networks …

Quantum computers could run more quickly if they contained tiny quantum refrigerators.

Quantum computers process information using quantum bits, or qubits, which must be reset to a special state between running programs. Teruaki Yoshioka at the Tokyo University of Science says slow reset times can become a real computational bottleneck, especially if they are slower than the speed of operations of a quantum computer.

He and his colleagues have devised a way to make this resetting process shorter by using a nano-sized quantum circuit refrigerator …

A computer that uses heat instead of electricity could run algorithms that power neural networks and artificial intelligence – and tamp down their energy budgets.

“We have things like ChatGPT which can learn very complicated things about language, but it consumes an amount of energy that is absolutely crazy,” says Nicolas Brunner at the University of Geneva in Switzerland. Some estimates put ChatGPT’s daily energy consumption on par with more than 30,000 households in the US.

Most modern AI technology …