Researchers now claim to have replicated the performance of Google's Sycamore quantum computer using traditional hardware.
Google claimed in 2019 its Sycamore system achieved the coveted quantum supremacy level of computing with its Sycamore system. The problem used to solve the claim has just been solved by today’s power-based accelerator. According to the report by Science, researchers in China recently solved the same computational problem that led Google to claim the title, but still being equipped with 512 GPUs, which were enough to accelerate the original algorithm resuscitation. The quantum theory, however, refers to the moment when a quantum computer is able to solve a problem that cannot be solved by a classical computer.
At that time Google said it would take the most powerful supercomputer – the IBM-provided summit – an unholy 10,000 years to solve the same computation that its quantum computer crunched in 200 seconds. It took fifteen hours to do that same thing with 512 GPUs.
Google’s claim to the quantum supremacy claim reposed on discovery of a semblance of interference in the qubit’s values. The quantum computing is a biffy master, and all of the current approaches to it are prone to decoherence. This means that the environment and the qubits design and operation make mistakes in computation.
After adjusting this system’s operations and running the same algorithm over Sycamore for 200 seconds (and millions of iterations), Google then extrapolated a result with the result showing the pattern of the processor’s departures from the exact, correct values that it should be outputting. These deviations occurred because of the mistake that increased the likelihood of certain outputs compared to others; finally, this pattern was visualized using a spiky graph that hardly ever reproduced.
This graphical representation of the relationship between errors and outputs is what Google claims that gave it the quantum top priority. That same graph was achieved by Chinese scientists. To accomplish this, they represented the problem through an interfering 3D mathematics array – a matrix – which enabled their 512 GPU’s advanced tensor cores to solve it by simply multiplying the values in the array.
But the engineers, which emphasized, pointed out one point: technology is continuously evolving and quantum computing is now beginning to run through the leaps and bounds phase that is now a few and far between occurrence for classical systems. Google’s Sycamore would be able to provide the spiky outline more fidelity than it did, at 2 p.m. – today’s quantum computers would be able to do it better, due to the improvements in errors correction.
The low fidelity achieved by Sycamore was the perfect bit to give to the Chinese scientists some extra space to just improve their calculations’ fidelity to 0.37%. That’s enough to beat Sycamore, but still a far cry from what is theoretically possible.
And while that is, too, likely to be true, it might be quite possible.