|
Project Topic
|
Quantum computers offer the potential to revolutionise our world. To realise this potential, however, we must overcome many challenges, from deep foundational issues in quantum information theory through to the extreme technical demands of high-tech engineering.
On the theoretical side, one of the longest standing open questions is to understand exactly which quantum resources are relevant for the advantages afforded by quantum computation. This is a roadblock in designing algorithms and applications that can fully exploit the power of quantum computing. Without knowing this, we can only ever be rather imprecise in what we are ultimately asking the engineers to create. In ResourceQ there are two key aspects to this problem that we seek to address. The first is that, to date, many seemingly disparate resources have been identified as being responsible for the quantum speedup, and which of these is actually relevant seems to depend on which model of computation one considers. We seek to understand whether or not there is any underlying resource which underpins all of these, that is, to find a model independent resource which is responsible for the speedup. The second is that there is a large gulf between the kinds of resources which are considered fundamental quantum resources and those that are typically relevant in a given experimental implementation. For example, nonlocality and contextuality are often considered the most fundamental signatures of the nonclassicality of nature, but (at least for near term devices) the more relevant resources are typically things such as quantum volume, cluster states, the number of T gates, or the depth of the quantum circuit. We aim to bridge this gulf by understanding how these fundamental and practical resources relate to one another. Key tools for ResourceQ will be drawn from, and contribute towards the development of: (i) structural methods (for example graphical languages) coming from logic and theoretical computer science that are well suited to addressing considerations like compositionality that we know to be at the heart of the differences between quantum and classical theories and phenomena; (ii) machine learning techniques, in particular, reinforcement learning, that are well suited to the allocation and optimisation of complex quantities such as quantum resources; and (iii) general methods from quantum information theory, in particular, the study of quantum resource theories. ResourceQ will be grounded through two industrial partnerships. On the one hand, working with Quandela will provide a focus on photonic technologies for quantum computation, where much less is known about the underlying quantum resources. On the other hand, our advisors at TWT will guide us towards a particular use case, that is, the optimisation of resources for the simulation of complex structures and materials via partial differential equations.
|
