Quantum singular value decomposer

Quantum computing is emerging as a new computational paradigm that promises substantial speed-ups (exponential in some cases [1]) for a variety of problems, with a potentiallylarge impact on science, knowledge and society. Quantum computers will harness the strange-looking laws of Quantum Mechanics to accelerate many computations by orders of magnitude [2]. However, they will require error-correction protocols before many important algorithms can be run on these machines. These will demand a large number of highquality qubits, which imply a much longer time of arrival for this technology, if it is to be “fully-fledged” and incorporate these schemes. For this reason, researchers have begun to heavily investigate about possible uses of the first generations of quantum computers that will not incorporate error correction but that nonetheless may surpass the capabilities of even the largest classical supercomputers in the world. These computers have been termed Noisy Intermediate-Scale Quantum (NISQ) computers [3]. A particularly appealing choice of quantum programs to be run on these machines are the so-called variational quantum circuits, or Quantum Neural Networks (QNN). These are hybrid algorithms that will coordinate both a classical and a quantum computer to preform a computation, and need to be specifically designed for each application