The calculation of the ground state energy of beryllium hydride (BeH₂) at the IBM TJ Watson Research Center has set the record for the largest molecular simulation performed on a quantum computer. The results, recently published in Nature, demonstrate that some important experimental landmarks on the path toward a universal quantum computer may not be too far in the future.

Solving the ground state energy of a system, such as a molecule or a quantum magnet, relies on calculating the interactions between its components, such as electrons or spins. As the number of components increase, the size of the system scales exponentially. The time required to calculate the solution to a large system without approximating particle-particle interactions on classical supercomputers is longer than the age of the universe. “Encoding that problem on a quantum computer could lead to a speed-up with respect to the exponential overhead required on classical computers,” says Antonio Mezzacapo, co-first author of the Nature publication.

Quantum computers present a more powerful alternative to classical computing. As opposed to classical computers, whose bits can only have values of 0 or 1, quantum computers run on qubits. Qubits, in addition to being in either 0 or 1, can also be a superposition of the 0 and 1 states. This distinction makes quantum computers much more efficient than classical computers at solving certain algorithms. Theoretical and experimental physicists are figuring out how to encode interesting problems onto existing quantum computer hardware.

“Our focus [for this work] was how can we extract the maximal quantum computational power from existing quantum hardware,” says Abhinav Kandala, the study’s other co-first author. Solving for the ground state energy in molecular systems requires evaluating the system’s energy for iterations of quantum trial states until a minimum is reached. To encode the trial state on a quantum computer, Mezzacapo and Kandala’s team devised an experimental strategy that efficiently chose the number of qubits and entangling operations to represent the interacting electrons in the molecular system. This calculation used a state-of-the-art quantum computer processor, whose initialization, operation, and measurement of qubits were accomplished before the laboratory environment disrupted the quantum state of its qubits, which occurred within tens of microseconds.

This study “represented the largest molecular simulation on quantum hardware to date,” Kandala says. To solve for the ground state energy of beryllium hydride, the team encoded the interactions between eight electron orbitals of the molecule onto six entangled qubits. The team also demonstrated the versatility of the algorithm and hardware by separately simulating a quantum magnetic system, where “the importance of entanglement [between the magnetic spins] is evident,” Mezzacapo says. The energies calculated by the quantum computer for the two systems were in agreement with numerical simulations of the experiment on a classical computer. 

These results generated lots of excitement in the quantum computing community. Theoreticians have predicted ways to implement sophisticated calculations on quantum computers, but “[it’s] only been in the last year or two that quantum hardware has been built up to the level where initial simulations can be run,” says Britton Plourde of Syracuse University. Plourde was not involved in the study published in Nature.

The final goal, of course, is to develop a universal quantum computer that can efficiently solve many different computational problems. This study is an important benchmark toward those efforts by revealing the size of some of the errors generated. To build a universal quantum computer, error correction schemes must be employed—coupling in additional qubits, also known as expanding the quantum volume, to check the calculation being run. According Kandala, “We’re looking at improving the quantum volume of our devices, but also at furthering our understanding of the errors that exist [for current quantum hardware systems].”

Read the article in Nature.