The primary case of running a simulation model that was not possible with a supercomputer with a quantum computer got here out. It was also evaluated that it will be a vital opportunity to prove the excellence of quantum computers.

On the 14th (local time), IBM, in collaboration with the UC Berkeley research team, proved for the primary time that a quantum computer can produce accurate results that exceed conventional computing methods on a scale of 100 qubits or more, and published a related paper in Nature. .

In quantum mechanics, which ends up in the microscopic world, matter can exist in several overlapping states, not one, so the computational ability increases remarkably. Existing computers express the presence or absence of electrons in units of 0 and 1, that’s, 1 bit. Compared, the unit of a quantum computer is a qubit by which states 0 and 1 are superimposed.

If a traditional computer has 2 bits, it is going to be considered one of 00, 01, 10, 11, but 2 qubits can do all 4 at the identical time. If there are 300 qubits, 2 to the facility of 300 states are possible, which is greater than the variety of all atoms within the universe, greatly increasing computer power.

Nevertheless, today’s quantum systems are inherently noisy and introduce many performance-destroying errors. That is because of the delicate nature of quantum bits or qubits and disturbances from their surroundings.

On this experiment, the IBM researchers first analyzed the characteristics of errors present in quantum computers and modeled materials with magnetic properties. To simulate the dynamics of ‘spin’, a fundamental component of magnetic materials, large-scale ‘entanglement’ states are created on the IBM ‘Eagle’ quantum processor made up of 127 superconducting qubits to calculate the behavior of magnetic materials and eliminate errors. It is claimed to have been supplemented.

The IBM researchers compared the performance of quantum computers with supercomputers at Lawrence Berkeley National Laboratory and Purdue University supercomputers. The issue was so complex that even supercomputers couldn’t calculate an actual answer. Then again, IBM’s quantum computer using 127 qubits took lower than one thousandth of a second to finish the calculation, the researchers said.

Previously, in 2019, Google solved an issue that might take a supercomputer 10,000 years in 3 minutes with a quantum computer of only 53 qubits. But on the time, quantum computers weren’t solving practical problems. Then again, IBM has succeeded in calculating the behavior of magnetic materials in its quantum processor Eagle.

“That is the primary time that a quantum computer has gone beyond conventional approaches and accurately calculated the physical phenomena of nature,” said Dario Gil, senior vp of research at IBM, on the day. ”he said.

Nevertheless, the outcomes of this study don’t mean that quantum computers have achieved “quantum supremacy,” which outperforms supercomputers. Nevertheless, this study is meaningful in that it showed that a quantum computer could make reliable estimations within the presence of errors.

Reporter Park Chan cpark@aitimes.com