Google is expanding the scope of its quantum computing ambitions by teaming up with a group of university researchers to design and build its own quantum processors. The company announced its plans in a blog post on Tuesday.
The researchers come from the University of California, Santa Barbara, and are led by John Martinis. They have invented a new type of quantum computing array they claim is more reliable than previous attempts to build such systems. According to an April press release describing their research:
“The unique configuration of the group’s array results from the flexibility of geometry at the superconductive level, which allowed the scientists to create cross-shaped qubits they named Xmons. Superconductivity results when certain materials are cooled to a critical level that removes electrical resistance and eliminates magnetic fields. The team chose to place five Xmons in a single row, with each qubit talking to its nearest neighbor, a simple but effective arrangement.”
Despite the new technologies Google’s Quantum Artificial Intelligence team is bringing on board, and its stated mission to design its own hardware, Google isn’t severing its ties with D-Wave Systems, the Canadian quantum computer manufacturer from which Google and NASA jointly purchased a system in 2013. In fact, Google’s Hartmut Neven says in the blog post, “We will continue to collaborate with D-Wave scientists and to experiment with the ‘Vesuvius’ machine at NASA Ames which will be upgraded to a 1,000 qubit ‘Washington’ processor.”
D-Wave’s unique brand of quantum computing has been a lightning rod of debate among scientists trying to prove and disprove its quantum nature, but the company has held fast to its claim that it has indeed built a commercially viable quantum computer. At our Structure Data conference in March, D-Wave CEO Vern Brownell noted that Google has successfully tested some new types of machine learning and computer vision algorithms on its D-Wave computer.
Still, it’s not surprising that Google would be interested in designing its own quantum computers if it could pull it off. The company has long designed its own servers and switches, and is pushing an artificial intelligence agenda that includes smartphones, robots and driverless cars. If Google, or anyone, is going to solve the very hard AI problems these technologies present, they probably can’t sit around and wait for someone else to build the right systems for them.