A team of scientists in China claims to have replicated the performance of Google’s Sycamore quantum computer using traditional hardware, undermining the suggestion that the company has achieved quantum supremacy.
As reported by Science magazine, the scientists used a system made up of 512 GPUs to complete the same computation developed by Google to show it had reached the quantum domination milestone in 2019.
The pursuit was led by statistical physicist Pan Zhang, who said his team’s supercomputer performed the calculation 10 billion times faster than Google thought possible.
Quest for quantum supremacy
Quantum supremacy (or quantum advantage) can be defined as the point at which quantum computers can exceed the maximum potential performance of classical supercomputers in a particular discipline.
Three years ago, Google announced that it had achieved this feat with Sycamore, which it said took just 200 seconds to solve a statistical math problem that would take a supercomputer 10,000 years to solve.
The problem was designed to highlight both the features of quantum computers, which use a phenomenon known as superposition to speed up computation, and the limitations of traditional systems.
However, the claim was: treated with suspicion by industry players from the outset. The loudest critic was IBM, who claimed that its Summit supercomputer would be able to solve the same puzzle in two and a half days, if given enough time to perfect the implementation. And now, a paper published by Zhang and his colleagues has further questioned Google’s claims.
As explained by ScienceTo replicate Sycamore’s performance, the scientists first had to reconfigure the problem as a 3D mathematical array called a tensor network. This meant they could use the system’s GPUs to run computations in parallel.
The low accuracy of Sycamore’s calculation gave the Chinese team additional leeway. By minimizing the accuracy target, they were able to speed up the calculation hundreds of times, replicating the same level of performance that the quantum machine had achieved.
While it appears that Google has been hasty in its celebrations, the quantum supremacy milestone is expected to fall sooner rather than later due to the level of investment spent on space R&D activities.
Efforts are underway to improve performance by increasing the number of qubits on a quantum processor, but researchers are also exploring possibilities related to qubit design, linking quantum computers and classical computers, new cooling techniques, and more.
Recently, IBM also suggested that: focus investments on error mitigation could bear significant fruit at this stage in the development of quantum systems and pave the way to quantum supremacy. The ability to dampen certain sources of error will make it possible to run increasingly complex quantum workloads with reliable results, the company says.
“At IBM Quantum, we plan to continue developing our hardware and software with this path in mind,” the company wrote in a statement. blog post.
“At the same time, together with our partners and the growing quantum community, we will continue to expand the list of problems we can assign to quantum circuits and develop better ways to compare quantum circuit approaches to traditional classical methods to determine whether a problem can be solved. . demonstrate quantum advantage. We fully expect that this ongoing path we have outlined will bring us practical quantum computing.”
- The advance of quantum computers requires a whole load of the best coding technologies
Through science (opens in new tab)