The name of the base system (platform): | Google TensorFlow |
Developers: | |
Date of the premiere of the system: | March, 2020 |
Technology: | Robotics, Development tools of applications, Supercomputer |
2020: Release of the solution to the market
In March, 2020 Google issued the platform open source of TensorFlow Quantum which allows developers to create AI models for quantum computers.
This expansion for TensorFlow offers set of operators — the low-level construction blocks in programming creating models of machine learning which work with qubits, quantum logic gates and quantum schemes. These operators undertake some difficult components to reduce code amount which should be written programmers.
TensorFlow Quantum allows researchers to build quantum data sets, quantum models and classical parameters of management as tensors in one computing graph — researchers of Google Alan Huo and Masoud Mokhseni write in the blog. |
By means of this framework it is possible to create hybrid and classical models of machine learning, to train diskriminativny and generative quantum models and to support emulators of quantum schemes.
One of potential scenarios of use TensorFlow Quantum is interpretation of quantum data. As for qubit it is possible to set at the same time values 1 and 0, search of results of the calculations executed by the quantum processor in itself is a serious problem. According to Huo and Mokhseni from Google, TensorFlow Quantum can help engineers to develop models of artificial intelligence which automatically untangle quantum data.
Also the companies have similar products Microsoft and IBM — Azure Quantum and IBM Q respectively.
By March 11, 2020 work of TensorFlow Quantum is limited to use of the imitated quantum computers, but Google is going to make in the future the solution compatible to the real systems.[1]