Hello, I'm Minato from blueqat, a buzzword-filled company aiming to become a quantum deep learning company with semiconductor quantum computers.
In the year 2022, the quantum computer industry is at a crossroads: the error-prone quantum computer known as NISQ is not useful, and the error-correcting FTQC has a long way to go. This is why we are aiming to become a quantum deep learning company. Let me explain what it is.
Development of Quantum Deep Learning
Quantum machine learning has made dramatic progress in the past few years. In particular, quantum machine learning using angle as a parameter, called variational circuits, has made great progress. In addition, simulation techniques for classical computers have been developed considerably to realize supercomputer calculations that can surpass the performance of quantum computers. Quantum machine learning is not necessarily just about using quantum computers, but also about using all the knowledge and technology gained in the process of developing quantum computers to facilitate machine learning. Quantum machine learning has been further developed, and theoretically deep quantum circuits are beginning to emerge for the depth direction. Basically, it is becoming clear that expressive power can be extended by using quantum entanglement or deep quantum circuits. Especially for data and models.
Quantum deep learning is hardware agnostic.
Quantum machine learning and quantum deep learning are not limited to quantum computers. It requires a well-balanced use of classical and quantum computers, quantum libraries and tensor libraries, and deep learning libraries for classical computers. In addition to the CPUs and GPUs available today, various NPUs can also be used. Many areas of HPC will be important. Quantum computers can also be used in conjunction with continuous and discrete quantity optical quantum computers, in addition to superconducting, ion trap, and other types of cold atoms, silicon, and diamond. All of these hardware types need to be well balanced.
Solving social problems, quantum problems, and realizing useful applications
Quantum deep learning can be applied to a wide range of data, from quantum data, which is data about the behavior of the quantum world, to data about our own world. Even if we are dealing with classical problems, we can devise models and other techniques to improve the efficiency of real-world problems. New ways of handling data can be realized that will contribute to a carbon-neutral society in the next generation with low power consumption. We can use data to make many useful social problems better.
The year 2022 is the perfect year to start working towards industrialization and business, including quantum deep learning. Let's make the world a better place with quantum and physics technologies.