Google has used its record-breaking quantum computer to complete the biggest quantum chemistry simulation ever, modelling the behaviour of a long chain of hydrogen atoms.
The simulation of large molecules is one of the main expected applications of quantum computers in the near future. Such molecules are difficult to fully understand using classical computers 鈥� the largest we can simulate accurately is pentacene, which contains 22 carbon atoms and 14 hydrogen atoms, says Jamie Garcia at IBM. 鈥淎s molecules get larger, they very quickly get out of the realm of what you can simulate with classical computers.鈥�
Quantum computers use quantum bits, or qubits, that exploit the properties of quantum physics to perform calculations. 鈥淭he atoms are quantum, the computer is quantum, we鈥檙e using quantum to simulate quantum,鈥� says Linghua Zhu at Virginia Tech, who was not involved in the work.
Advertisement
鈥淲hen we use classical methods we always use approximations, but with a quantum computer it鈥檚 possible to exactly know how each atom is interacting with the others,鈥� says Zhu.
Until now, the largest molecule simulated with a quantum computer was beryllium hydride, which is one beryllium atom and two hydrogen atoms, says Zhu.
But at the Q2B conference in California on 10 December, Google鈥檚 head of quantum algorithms, Ryan Babbush, announced his team has managed to simulate a much larger molecule 鈥� a chain of 12 hydrogen atoms.
鈥淭hat鈥檚 a result that we鈥檙e pretty excited about, because this is more than double the number of qubits and the number of electrons as any prior quantum chemistry simulation, and it had the same level of accuracy,鈥� Babbush said.
The team used the same Sycamore quantum computer recently used to achieve quantum supremacy 鈥� a calculation not possible on an ordinary computer 鈥� but didn鈥檛 use all of the qubits for the chemistry simulation, he said.
While this type of hydrogen molecule doesn鈥檛 actually exist in nature, chains of hydrogen atoms are commonly simulated in quantum chemistry because they are relatively simple systems that are easy to add to, Babbush said.
This isn鈥檛 a calculation that would be impossible for the best classical computers, but it demonstrates the strength of Google鈥檚 quantum computer, Zhu says.
