Nvidia will help Google develop efficient quantum processors

Nvidia will help Alphabet, Google’s parent company, develop quantum processors. Google’s Quantum AI division will use Nvidia’s Eos supercomputer to speed up the design of quantum components, according to a statement from both companies.

Image source: NVIDIA

The idea is to use Nvidia’s Eos supercomputer to simulate the physical processors needed to run quantum processors, which will help overcome current limitations in developing truly efficient quantum systems.

Quantum computing is based on the principles of using quantum mechanics to create machines that are much faster than current semiconductor-based technologies. However, the time has not yet come for the widespread use of such technologies. As Bloomberg reports, while various companies have announced breakthroughs in quantum computing, it may take decades for truly large-scale commercial quantum computing projects to hit the market.

Nvidia, the world’s most valuable company, believes its hardware technology will help Google solve one difficult problem related to quantum computing. As quantum processors become more complex and powerful, quantum computing becomes increasingly difficult to distinguish between actual information and interference known as noise.

«Developing commercially useful quantum computers is only possible if we can scale quantum hardware while controlling noise. Using Nvidia’s accelerated computing, we are studying the impact of noise on the increasing complexity of quantum chip circuitry,” said Guifre Vidal, Google Quantum AI Research Scientist.

To find solutions, Nvidia proposes using a giant supercomputer that uses its AI accelerators. With the help of a supercomputer, the processes of interaction of quantum systems with the environment will be simulated. For example, many quantum chips need to be cooled to very low temperatures in order for them to work at all.

Previously, such calculations were extremely expensive and time-consuming. Nvidia says its system will produce results from calculations that would previously have taken a week in minutes, at a significantly lower cost.