Google's new quantum computer, Willow, surprised the world just two days after its unveiling. Alphabet, Google's parent company, saw its stock rise by more than 5% in the U.S. stock market.
On Sept. 9 (local time), Google announced in the journal Nature that the new quantum computer Willow solved a problem that would take 10 septillion years with existing supercomputers in just 5 minutes. Five years ago, in 2019, Google introduced the Sycamore quantum computer, proving for the first time that quantum computers outperform supercomputers in what is referred to as 'quantum supremacy.' However, as the performance of supercomputers improved, quantum supremacy became a thing of the past, and now, Willow has demonstrated performance that far surpasses supercomputers again after five years.
Julian Kelly, a physicist at Google Quantum AI who participated in the development of Willow, explained, 'We maintained all the advantages of Sycamore while modifying the internal structure and reconfiguring the processor.' Sycamore had 67 qubits, whereas Willow has increased this to 105.
Google chose the Frontier supercomputer to compare Willow's performance. Frontier, located at the Oak Ridge National Laboratory, is the world's first exascale supercomputer. While it has slipped to second place this year after the newly released El Capitan supercomputer at Lawrence Livermore National Laboratory in the U.S., it remains one of the most dominant supercomputers in the world.
Google researchers presented a benchmark task called Random Circuit Sampling (RCS) to both Frontier and Willow. This task validates how close the sequence of outputs from programs run on the chip is to being random, checking the computer's performance. Frontier was found to take 10 septillion years to complete this task. Ten septillion years is equivalent to 10 years in Korean. The age of the entire universe is still less than 10 septillion years. In contrast, Willow solved the task in just 5 minutes.
The key to Willow's overwhelming performance lies in 'accuracy' rather than 'size.' It is often thought that the performance of quantum computers is proportional to the number of qubits. A qubit is the basic unit of information processing in a quantum computer, capable of processing information in both 0s and 1s simultaneously. Various corporations and countries around the world are striving to build quantum computers with larger numbers of qubits. Willow employs 105 qubits, which is now quite common. IBM has created a 1,000-qubit quantum computer called Condor, and a 127-qubit quantum computer is currently operational at Yonsei University's Songdo campus.
Instead of increasing the number of qubits, Google researchers focused on enhancing the quality of each individual qubit and reducing the error rate. While qubits can process information in both 0s and 1s, they also come with a higher probability of errors. As the number of qubits increases, the probability of errors rises as well, making it essential to lower the error rate as a significant task for the commercialization of quantum computers.
Julian Kelly from Google explained, 'Willow's qubits can maintain quantum states more than five times more accurately than Sycamore, allowing stable encoding of information.' Google researchers adjusted the qubits' configuration to ensure that the error rate would not increase even as the number of qubits grew, successfully halving the error rate even when the number of qubits was doubled. They grouped multiple qubits to create a single 'logical qubit,' using this logical qubit to reduce errors.
Martin Weides, a professor at the University of Glasgow, noted, 'This research presents directions for creating quantum computers that eliminate quantum errors,' adding, 'There are still challenges to overcome, but this progress is significant.'
Google plans to embark on research that solves real-world problems rather than just benchmark tasks using Willow. For the commercialization of quantum computers, it is essential to demonstrate that they can be utilized in real-life applications beyond simple benchmark performance. Julian Kelly stated, 'New challenges for all fundamental science and engineering related to quantum computers are emerging.'
The competition for technological development related to the commercialization of quantum computers is intensifying. South Korea has also designated quantum technology as a top-three game changer and has increased support, but recent developments regarding the impeachment of President Yoon Suk Yeol have created a zero-sum situation. The government initially planned to launch a Quantum Strategy Committee chaired by the Prime Minister by the end of this month or early next month to support the development of quantum technology and the nurturing of the industry. However, due to the uncertainty surrounding the positions of committee members such as the Prime Minister, Minister of National Defense, and Director of National Intelligence amid the impeachment situation of President Yoon, it remains unclear whether the committee can be established normally.
A researcher in the quantum field stated, 'President Yoon Suk Yeol's government has announced active investment plans for quantum technology, but over time, there seems to be a drifting away from interest, and support is not being implemented properly,' adding, 'There is not much time left to avoid falling behind in the technological competition with global companies like Google or IBM.'