Computational power for solving intractable problems
Computer power has increased dramatically over the last decades by miniaturization of the active components dictated by Moore’s law. Because device dimensions are approaching the atomic scale, it is expected that further downscaling will stop within the coming decade. Many applications, such as artificial intelligence (AI), require more powerful computers, though, which, at the same time, should be more energy-efficient. Therefore, alternative computational concepts, with different scaling laws, should be explored. Quantum and neuromorphic computation are paradigms that can achieve that.
Quantum computing has the potential to solve problems that are practically impossible for classical computers. It makes use of quantum mechanical effects, with superposition and entanglement of the qubits playing an important role. This results in a fundamentally different way of operation, giving rise to an exponential scaling of computing power with the number of qubits. That opens the way for revolutionary applications such as the resolution of complex optimization challenges, or the prediction, simulation and modelling of the behavior of molecules, catalysts and new materials. Neuromorphic computing is inspired by the architecture of the brain and it is generally believed that it can offer great opportunities in complex classification and AI tasks like machine learning, and at the same time be extremely energy-efficient.