Google announced in 2019 that they had defeated the world’s most powerful supercomputer with a quantum computer called Sycamore. New research by Chinese scientists showed that GPUs used in normal computers can beat Sycamore with the computing power.
Google’s quantum computer, named Sycamore, surpassed Summit, the most powerful supercomputer of the time, in solving the same problem. According to Google, it was announced that Summit would solve the problem in 10,000 years. Sycamore solved the same problem in just 200 seconds. So this was an indication that Google had achieved quantum supremacy.
What is quantum supremacy?
Quantum supremacy is a term first used by John Preskill. Quantum supremacy is the ability to solve a problem that no classical computer can solve in practice, by a quantum computer in a practical time. In quantum supremacy, it is sufficient to find a better job than classical computers.
The 10,000-year period revealed in Google’s test means that the problem cannot be solved in practice. Therefore, as a result of this test, it is accepted that Google has achieved quantum supremacy.
Google’s quantum supremacy isn’t that superior
In the research conducted by Chinese scientists, the original problem was solved thanks to the processing power of today’s GPUs, with the support of advanced algorithms. According to the article published in Arxiv, the research used 512 GPUs for mathematical operations, while advanced algorithms were used to solve the problem. As a result of all these, the problem was solved in 15 hours. However, according to the researchers, if they had a supercomputer like Summit, they could probably solve this problem in just a few seconds.
Google says the future is in quantum computing
The working principle of quantum computers and classical computers is completely different. Because a classical computer tries to reach the result by handling the operations in order. For example, trying to find the password of a file; While classical computers try all the variations one by one and reach the result, quantum computers try all the variations at the same time. However, this does not prove that every transaction will be faster in quantum computers.
Referring to the research, Google announced that they have already said in their 2019 articles that classical algorithms will develop. Technology giant Google said, “Despite significant developments over the past few years, we do not think the classical approach will be able to keep up with quantum circuits in 2022 and beyond.” made a statement. Google’s Sycamore quantum computer was a 53-qubit processor, while the IBM Eagle, introduced in 2021, has 127 qubits. According to IBM’s roadmap, it is predicted that the 1000 qubit threshold will be broken in 2023. Therefore, Google’s assumption may not be considered unjustified.
13685
