Google has partnered with UCSB (University of California, Santa Barbara) to build quantum processors designed for applications in the field of artificial intelligence. In principle it is a new research project for this purpose.
Hartmut Neven, Google's chief engineering officer, announced the initiative via Google's Research Blog (Hardware Initiative at Quantum Artificial Intelligence Lab). The team, led by physicist John Martinis of the University of California at Santa Barbara (UCSB), will research and develop new quantum information processors based on electronic superconductors with the aim of expanding artificial intelligence technologies.
The Google executive noted that Martinis and his team at UCSB have made "great strides" in building quantum superconducting electronic components, and the researcher was also recently awarded the London Prize for his "pioneering advances in quantum control and processing. of quantum information ”.
|John Martinis (UCSB), London Prize|
The new team will be located in the Quantum Artificial Intelligence Lab, with a collaborative effort between Google, NASA Ames Research Center, and the Universities Space Research Association (USRA).
The project will allow the implementation and testing of new designs of quantum inference and optimization processors based on recent theoretical knowledge, as well as lessons learned from the well-known D-Wave quantum architecture. "
Although Google will determine the team to build its own quantum processor designs, the company maintains that it will continue to collaborate with scientists at D-Wave and to experiment with the "Vesuvius" machine at NASA.
|NASA Support structure for the installation of the D-Wave Vesuvius processor,|
which cools down to 20 millikelvin (close to absolute zero).
While standard computers handle binary data - which is expressed in zeros and ones - quantum computing relies on the behavior of subatomic particles.
That is, quantum computing is based on quantum bits or qubits. Unlike traditional computers, in which bits must have a value of zero or one, a qubit can represent zero, one, or both values simultaneously. The representation of information in qubits allows information to be processed in ways that have no equivalent in classical computing, which take advantage of phenomena such as the quantum tunneling effect and quantum entanglement. As such, quantum computers could theoretically be able to solve certain problems in a couple of days, while a classical computer would take millions of years.
Some theorists believe that qubits could greatly improve the speed and power of computing.
Going beyond self-driving cars, balloons, and autonomous Wi-Fi robots, Google has shown a growing interest in artificial intelligence in recent years. In January, the tech giant acquired British artificial intelligence firm Deepmind for what is believed to have paid $ 400 million.
Charlie Osborne http://www.zdnet.com/google-launches-quantum-processor-artificial-intelligence-project-7000033241/