A SUPERIOR quantum computer might break encryption and tackle issues that conventional computers couldn’t. Although no one has yet succeeded in creating such a gadget, research has lately accelerated — may 2022 be the year?
Efforts are now focused on achieving quantum supremacy, which is the point at which a quantum computer can execute a computation that a conventional computer cannot in an acceptable period of time.
Google was the first to achieve this aim in 2019, utilizing a gadget with 54 qubits, the quantum counterpart of ordinary computer bits, to do a random sampling calculation, which is practically meaningless. A team from China’s University of Science and Technology solved a more sophisticated sampling issue with 56 qubits in 2021, and then went much farther with 60 qubits in 2022.
However, IBM’s Bob Sutor believes that this game of leapfrog is mostly an academic achievement that has yet to make a significant influence. True supremacy will only be attained when a quantum computer outperforms conventional computers and can solve a variety of problems, rather than the random sampling computations that are now employed as benchmarks.
According to him, IBM is aiming toward achieving “quantum commercial advantage,” or the moment at which a quantum computer can solve really valuable issues for academics or businesses much quicker than traditional computers. According to Sutor, this hasn’t come yet and won’t till 2022, but it’ll be here within the decade.
More bullish is Nir Minerbi, co-founder of quantum software firm Classiq. He predicts that proof of quantum supremacy in a relevant problem will occur in 2022.
“Do you recall when the first electric automobiles were released? They were convenient for going to the store, but not so much for driving 300 miles to drop your child off at college. Quantum computers, like electric automobiles, will improve with time, making them usable in a wider range of applications,” he argues.
Solving practical issues has a lot of stumbling blocks. The first is that devices will require thousands more qubits, which must be more robust and trustworthy than current ones. To function as a single “logical qubit,” researchers will most likely need to arrange them together in clumps. This improves fidelity, but it eats into scale gains: hundreds of logical qubits may need millions of physical qubits.
Quantum error correction is also being developed by researchers to rectify errors as they arise. In July 2021, Google revealed that their Sycamore processor could identify and correct faults in superconducting qubits, although the additional hardware required to do so caused more errors than it corrected. With their trapped-ion qubits, researchers at the Joint Quantum Institute in Maryland were able to break beyond that vital break-even point.
Even yet, it’s still early. It would be “quite stunning,” according to Scott Aaronson of the University of Texas at Austin if a general-purpose quantum computer solved a significant issue in 2022. “Error correction is just getting started, and we don’t appear to be anywhere near being able to safeguard a single encoded qubit for an arbitrary period of time, much alone executing calculations on thousands or millions of encoded qubits,” he adds.