Google, in collaboration with researchers from Harvard, Lawrence Berkeley National Labs, UC Santa Barbara, Tufts University and University College London, announced that they have achieved the first completely scalable quantum simulation of a molecule. Quantum-level chemistry simulation is one of the first real-world uses for quantum computers, and it could revolutionize medicine research, materials research, and much more.
Classical computers aren’t that good at simulating chemical reactions. For instance, accurately computing the energies of the propane molecule (C3H8) takes ten days with a classic computer design. Molecular systems use highly-entangled superposition states, which require exponentially more computing resources to represent them sufficiently with high precision.
If it takes up to ten days to represent a single molecule with high precision, then representing thousands (or millions) of molecules at the same time, and how they react with each other, becomes an almost impossible task with a classic computer architecture. It is also a highly inefficient process in terms of energy use. Chemical reactions seem to work at a quantum level, so quantum computers are ideal for trying to simulate them as efficiently and as accurately as possible.
Google has been researching quantum computers for many years, primarily since it started collaborating with D-Wave, a Canadian company. D-Wave doesn’t have a universal quantum computer, but its quantum-annealing computer specializes in solving optimization problems. Google has also been working on building a much more useful (but still currently limited) universal quantum computer with a few stable qubits, which the company seems to have used for this experiment.
Google was able to compute the energy landscape of the hydrogen H2 molecule using a quantum computer. To prove that it worked, the team behind the experiment compared the results with the results obtained from a classically computed simulation. They lined up almost perfectly.
Simulating much larger scale chemical systems would be impossible on traditional computers, but quantum computers may provide the solution to that problem. For instance, Google believes that when universal quantum computers achieve about one hundred qubits, it should be possible to simulate how bacteria produce fertilizer at room temperature.
Producing fertilizer requires 1 to 2 percent of the global energy production, according to Google, which means we haven’t discovered how to make it efficiently yet. Quantum computers may provide us these answers. Once quantum computers have enough qubits they could also aid in discovering more efficient solar panel cells, better batteries, materials, medicine, and so on.