Abstract
Computer-aided engineering simulations, in particular, Computational Fluid Dynamics, have become a fundamental design and analysis tool in product development. Over time, a demand for larger problem sizes and higher accuracy has led to huge computational workloads requiring extended compute capabilities. Increasing computing capabilities requirements, however, drive a fast-growing power consumption. In order to deal with increasing power demand, hardware and software solutions' reevaluation in terms of power-efficiency becomes of paramount importance. Establishing a power budget and reducing the compute units operating frequency in order to comply with such budget is becoming common practice. However, in the presence of heterogeneous compute units and dynamic workloads, a static and uniform reduction across compute units leads to a potentially severe impact on performance. This paper proposes a run-time heterogeneity-aware power-adaptive schedule that provides power consumption optimization, targeting heterogeneous parallel distributed systems in the context of CFD simulations. The proposed approach is integrated into OpenFOAM computational library and explores power migration and reduction across nodes, considering runtime workload imbalances and node performances. Results reveal not only a substantial reduction in power usage but also significant performance gains relative to the uniform static approach. To the best of authors' knowledge, this is the first implementation and integration of power management solutions in OpenFOAM.

Abstract
This extended abstract reports on on-going research on quantum algorithmic approaches to the problem of generalised tree search that may exhibit effective quantum speedup, even in the presence of non-constant branching factors. Two strategies are briefly summarised and current work outlined.

Abstract
Reinforcement Learning is at the core of a recent revolution in Artificial Intelligence. Simultaneously, we are witnessing the emergence of a new field: Quantum Machine Learning. In the context of these two major developments, this work addresses the interplay between Quantum Computing and Reinforcement Learning. Learning by interaction is possible in the quantum setting using the concept of oraculization of environments. The paper extends previous oracular instances to address more general stochastic environments. In this setting, we developed a novel quantum algorithm for near-optimal decision-making based on the Reinforcement Learning paradigm known as Sparse Sampling. The proposed algorithm exhibits a quadratic speedup compared to its classical counterpart. To the best of the authors' knowledge, this is the first quantum planning algorithm exhibiting a time complexity independent of the number of states of the environment, which makes it suitable for large state space environments, where planning is otherwise intractable.

Abstract
Bayesian Monte Carlo (BMC) is a promising integration technique which considerably broadens the theoretical tools that can be used to maximize and exploit the information produced by sampling, while keeping the fundamental property of data dimension independence of classical Monte Carlo (CMC). Moreover, BMC uses information that is ignored in the CMC method, such as the position of the samples and prior stochastic information about the integrand, which often leads to better integral estimates. Nevertheless, the use of BMC in computer graphics is still in an incipient phase and its application to more evolved and widely used rendering algorithms remains cumbersome. In this article we propose to apply BMC to a two-level adaptive sampling scheme for illumination integrals. We propose an efficient solution for the second level quadrature computation and show that the proposed method outperforms adaptive quasi-Monte Carlo in terms of image error and high frequency noise. © 2016 The Eurographics Association.

Abstract
This article proposes a Markov Chain Monte Carlo (MCMC) rendering algorithm based on a family of guided transition kernels. The kernels exploit properties of ensembles of light transport paths, which are distributed according to the lighting in the scene, and utilize this information to make informed decisions for guiding local path sampling. Critically, our approach does not require caching distributions in world space, saving time and memory, yet it is able to make guided sampling decisions based on whole paths. We show how this can be implemented efficiently by organizing the paths in each ensemble and designing transition kernels for MCMC rendering based on a carefully chosen subset of paths from the ensemble. This algorithm is easy to parallelize and leads to improvements in variance when rendering a variety of scenes.

Abstract