D-Wave’s assertion of quantum computer dominance is being doubted

D-Wave’s Advantage2 Quantum Computer Chip

Kent Kallberg/D-Wave

Quantum Computing Firm D-Wave says the device can solve problems that are virtually impossible with a classic computer, but two separate research groups have raised questions about the claim.

As reported by New Scientist Last year, D-Wave published a preprint claiming that its benefits quantum computers can calculate problems with transverse magnetic field ISING models. This is a quantum version of a mathematical approximation of the behavior of a substance when changing the state from liquid to gas.

The article has now passed peer reviews and is published in the journal Science At the same time as March 12th, researchers working on classical algorithms have shown that these problems are easily accessible on regular machines.

Dry the cell New York University and his colleagues say they have I did a similar calculation on a regular laptop in just 2 hoursuses a field of mathematics called tensor networks. These networks essentially reduce the amount of data that simulation requires, and significantly reduce the computational power required to run.

Andrew King At D-Wave, this says it does nothing to change the company's original claims. “They didn't do all the problems we did. They didn't do all the sizes we did. “So it's a huge advancement and these are great researchers…but it doesn't refute our claims of superiority.”

King says after hearing about the SELS paper, he decided to perform larger calculations that include up to 3200 qubits (quantum bits, building blocks of quantum computers). He says this shows even more quantum advantage, but the results have not been published yet.

SELS calls this response “a little get” and says his tensor approach can be easily scaled even further. The time to run the algorithm is scaled linearly proportional to the size of the problem, so there is no need to test larger problems. “If that's really going to make them [D-Wave] Super happy, and they'll say, “Okay, you guys did that,” we could do that,” Cels says. “I'm not planning. I can't see the points.”

In particular, Linda Mauron and Giuseppe Carreo Yokohata is facing a problem with the model in the EPFL of Lausanne, Switzerland. Solve without the need for quantum entanglement – A key source of the supposed benefits of quantum computers – or by simulating minimal amounts of entanglement on a regular computer.

Calleo says they are in a hurry to publish their papers so that the pair matches the d-wave Science Publications, and he acknowledges that it focuses only on one type of issue the company is addressing, not reaching the same size. D-Wave's paper suggests that such calculations take up to 200 years on a powerful supercomputer, Carleo says, but he and Mauron did it in three days using four graphics processing units (GPUs). That said, he says it should be possible to exceed the size of the problem solved by D-waves within a week.

“The lesson to learn is, “If you say, 'This goes beyond classic simulations,' then there's a classic simulation that does that,” Carleo says. “When they write these papers, my suggestion is to avoid these claims because they don't need them.”

In response, a D-Wave spokesman rejected these calculations. “In our paper, we found that this type of simulation is so easy that it cannot make a strong argument,” the spokesman says. “This paper appears to be progressing, but it does not challenge our argument beyond classical quantum simulations.”

If the results are confirmed to be overturned, it is not the first time a quantum computer has been touted as invincible, and it is only proven otherwise. In 2019, Google argued that the Sycamore Quantum Computer can perform calculations that take even the most powerful classic supercomputer in the world to 10,000 years. However, in 2022, the researchers completed the task in around 15 hours using 512 GPUs, and in 2024 another team completed the same task in 14.22 seconds. These classical speedups also relied on tensor networks.

Alex Kissinger At Oxford University, D-Wave was one of the first startups working on quantum computing, and in 2011 it provided what was called the first commercial quantum computer. However, the company was plagued by experts' questions about whether computers are truly quantum or whether they are excellent, unusual classic machines with specific optimization issues.

The question about quantum nature of D-Wave is more or less in bed at this point, but it remains to be seen whether the device can actually solve problems that are not possible with a regular machine. “Broadly, these days they seem more respected than they used to have made these big claims and kept all the details a bit under the hood,” says Kissinger. “Now, we can see quite a bit of detail about what their devices actually do.”