Quantum computers have the potential to solve problems much faster than classical computers. Google made a breakthrough in 2019 when it achieved quantum supremacy, performing a calculation that was impossible for the best classical computers at the time. However, the development of functional quantum computers is still facing challenges.
One of the factors that determine quantum supremacy is the number of qubits used and how they are programmed. A large enough number of qubits, combined with complex programming, can outperform classical computers. However, the exact threshold at which classical algorithms can no longer catch up is unclear.
Google’s original demonstration of quantum supremacy involved a task called random circuit sampling. They used 54 superconducting qubits for 20 cycles. More recently, researchers at Google performed the same feat with 70 qubits for 24 cycles. While this represents a jump in complexity, the quantum system is still plagued by “noise” that affects its performance and makes it vulnerable to classical advancements.
To measure the performance of quantum computers and their vulnerability to noise, researchers use benchmarks. These benchmarks involve comparing the outputs of a quantum machine with the predictions made by a classical computer. Google and other researchers have determined the level of noise at which this benchmark is still effective. This allows for a fair comparison between different generations of quantum computers.
Another group of researchers at the University of Science and Technology of China (USTC) has also demonstrated quantum advantage using a different type of quantum computer called Jiuzhang. This machine performs boson sampling, which measures the behavior of photons bouncing around a maze of mirrors and beam splitters. However, verifying the true quantum nature of the measurements is challenging and currently lacks a coherent method.
While progress is being made in the development of quantum computers, practical applications are still elusive. Researchers are exploring how quantum computers can be applied to solve real-world problems such as graph problems for drug design and machine learning. However, the verification problem and the potential for classical algorithms to achieve similar performance pose challenges.
Mapping real-world problems to quantum computers and vice versa will be a crucial part of research and development in the coming years. Scientists are working on applying quantum machines to various fields, including high-energy physics, materials, life sciences, and finance. The goal is to find specific problems where quantum computers can demonstrate clear advantages.
Ultimately, the true advancement of quantum computers may be defined not by benchmarks and mathematical proofs but by the adoption of these machines by scientists in various fields. When researchers outside the quantum information community start using quantum computers for their work, it will signal a significant milestone in the development of functional quantum computers.
