Google’s Sycamore breakthrough doesn’t spell the end for China’s hopes of winning quantum computer race
- After US tech giant unveiled a chip that dramatically outperformed supercomputers, Chinese researchers announced their own groundbreaking particle experiment
- Team still confident it can achieve what one scientist has described as the ‘holy grail’ of technology and say race between two countries is only just beginning
When Google’s scientists announced last month that they had achieved quantum supremacy – a breakthrough that could revolutionise the future of computing – Chinese scientists countered with news of their own breakthrough in the field.
Quantum computing may one day produce machines capable of performing calculations that are far quicker than any of today’s supercomputers can achieve, and the technology could transform – or disrupt – almost every aspect of society, from health care to finance to the military.
While Chinese scientists admit that the US is still the world’s leader in the field, they said the race to develop what one scientist described as the “holy grail” of science and technology was only just beginning.
The Chinese researchers from the University of Science and Technology of China in the city of Hefei said they had created a light-based simulation, which allowed them to make a calculation on scale 10 billion times greater than the best achieved using traditional methods.
The machine they developed will be the most advanced quantum computer that uses subatomic particles of light to perform calculations.
The announcement came a day after Google said that its groundbreaking Sycamore superconducting quantum chip had taken just 200 seconds to complete a task that the world’s most powerful computer would need 10,000 years to finish.
Their findings, an early version of which was accidentally published by Nasa and later published in full in the journal Nature, showed for the first time that a quantum computer could outperform traditional machines decisively on certain tasks.
The Chinese device, though advanced, is no match for Sycamore in terms of performance.
Professor Pan Jianwei, the leader of China’s quantum computing team, admitted that the US had outpaced China for now, but was adamant this is “definitely not the end”, while Huang Heliang, a member of Pan’s team, said the Chinese team still hoped to achieve quantum supremacy
Both achievements have helped China and the US take a step forward in quantum computing, which, by giving scientists control over subatomic particles, will one day lead to the development of much more powerful computers.
Possible breakthroughs will include life-extending medicine, machines with human-like intelligence and super-materials.
They could also have a disruptive impact on the existing order in financial markets, government, international community and military.
All this is possible because quantum computers will be able to perform computing tasks much more quickly than a regular computer.
For example, the technology could make mathematical encryption useless, so existing security measures such as passwords would no longer be able to protect bank accounts, government servers or military communications.
In the field of medicine the quantum computer’s greater efficiency in simulating complex physical or chemical processes would allow scientists to test things at a molecular level much more quickly, thereby aiding the development of new drugs.
Race picking up pace
While traditional computers store information in bits – effectively a string of binary code – quantum computers use quantum bits, or qubits, which can be set at 1 or 0 at the same time.
This means they can run a multitude of tasks simultaneously and perform calculations much more quickly.
The two achievements are also a sign that the race to build the first useful quantum computer is picking up pace – in part driven by the competition between China and the US to be the technological leader, according to scientists from both countries.
The Sycamore, for instance, has been regarded by some Chinese quantum scientists as a result of this competition. Before the Sycamore, US progress on quantum computers had been slow and had suffered many setbacks.
Two different types of machine built by the tech giant had failed – in part because simulations by Chinese scientists on Taihu Light, the nation's fastest supercomputer, showed that the older generations of Google quantum computers would never beat traditional machines unless they could expand their scale significantly, which was technically infeasible.
“It's obvious that there's an element of national competition in the interest of both countries' governments in this field,” according to Scott Aaronson, a professor of physics with the University of Texas at Austin.
“Roughly speaking, I would say that China is ahead of the US right now in quantum communications … whereas the US is ahead in quantum computation,” Aaronson added.
Focusing on quantum communications, a technology to protect transmitted data from being broken by a quantum computer, has been the primary focus of China’s quantum researchers.
Over the last decade China has built up the longest, most sophisticated quantum communication networks on Earth and launched Mozi, the world's first quantum satellite.
But this focus has left China lagging in other areas. While Google started up a quantum computer programme 13 years ago, and other American companies such as IBM have made intensive investments and built prototype machines, similar work did not begin in China until recently, according to quantum scientists.
But for now, a usable quantum computer is still a long way off, and neither the American nor Chinese advances have any real-life applications at present.
For example, none of the quantum computers is capable of performing relatively simple mathematical calculations and the current systems are purely to demonstrate the capabilities of the technology.
Pan’s team, for instance, built a quantum computer with sophisticated laser systems and used photons, or particles of light.
The team’s experiment, known as boson sampling, was conducted to calculate the likelihood of a particle leaving a network of beam splitters – a precision device used to split light – through a specific output portal.
When the light particle enters the network and hits a beam splitter, it can travel in different directions, depending on its quantum state.
The more particles there are, and the more layers of beam splitters they need to go through, the more difficult it is for a traditional computer to simulate the process because it involves some very sophisticated physical mechanisms.
At the end of the network there were numerous output portals serving as an exit point for the light particles. The quantum machine was able to make this calculation, akin to working out where a ball in a pinball machine would go – by simulating the movement of the particles in the network.
Using 20 photons as input, and operating in 3,600 optical modes, the Chinese quantum machine generated up to 100 trillion sets of possible configurations – or quantum state spaces – breaking three international records, according to the team.
A manuscript of their study, published on preprint server Arxiv.org, is under peer review.
However, Aaronson, an original designer of the boson sampling experiment, also said the Chinese light quantum computer had some serious limits.
The machine could observe only 14 out of the 20 input photons, which means its actual performance is only as powerful as a smartphone, according to Aaronson.
The difficulty of simulating the experiment on a classical computer is expected to grow by around two to the power of the number of observed photons.
“Two to the 14th power is 16,384 … this is still well within the range where a smartphone could easily simulate the experiment,” he said.
To demonstrate quantum supremacy with their light machine, Pan's team would need to increase their observed photons at least fourfold.
By way of comparison, Google's 53-qubit device can push the calculation challenge to the scale of nine quadrillion, large enough to severely tax the most powerful supercomputers on Earth.
But Wu Biao, a professor of quantum physics at Peking University, said the gap between China and US on quantum computer technology was not as big as this suggested.
“We haven’t left the starting line yet. If there is any difference, it may not matter very much because we are still far, far away from any real applications,” he said.
The results generated by quantum computers today are still highly unreliable. Sycamore, for instance, could only guarantee 0.2 per cent accuracy after 20 rounds of simulation.
And the particles in quantum computers are also easily affected by surrounding environments due to their fragile condition.
While the central processing unit of a laptop can operate for years without a glitch, the Google quantum chip has to stop every few microseconds.
Google's claims for its quantum chip have also been questioned by some in the US science community. IBM, for instance, has released a study alleging that Google's quantum simulation was exaggerated because it could be done on a classic computer in a couple of days, if not faster.
Nonetheless, the “quantum computer is the holy grail of human science and technology. It is drawing enormous interest from the government, private investors and the general public,” said Wu.
But high hopes for an immediate breakthrough are unrealistic, and may force quantum researchers “walking on a thin line between science and fiction” to seek more funding and support, he added.