Abstract

Abstract

Abstract
Rendering photorealistic images is a costly process which can take up to several days in the case of high quality images. In most cases, the task of sampling the incident radiance function to evaluate the illumination integral is responsible for an important share of the computation time. Therefore, to reach acceptable rendering times, the illumination integral must be evaluated using a limited set of samples. Such a restriction raises the question of how to obtain the most accurate approximation possible with such a limited set of samples. One must thus ensure that sampling produces the highest amount of information possible by carefully placing and weighting the limited set of samples. Furthermore, the integral evaluation should take into account not only the information brought by sampling but also possible information available prior to sampling, such as the integrand smoothness. This idea of sparse information and the need to fully exploit the little information available is present throughout this book. The presented methods correspond to the state-of-the-art solutions in computer graphics, and take into account information which had so far been underexploited (or even neglected) by the previous approaches. The intended audiences are Ph.D. students and researchers in the field of realistic image synthesis or global illumination algorithms, or any person with a solid background in graphics and numerical techniques. Table of Contents: Introduction / Spherical Fibonacci Point Sets for QMC Estimates of Illumination Integrals / Bayesian Monte Carlo for Global Illumination / Bibliography / Authors' Biographies Copyright © 2015 by Morgan & Claypool.

Abstract

Abstract
Exploiting the computing power of the diversity of resources available on heterogeneous systems is mandatory but a very challenging task. The diversity of architectures, execution models and programming tools, together with disjoint address spaces and different computing capabilities, raise a number of challenges that severely impact on application performance and programming productivity. This problem is further compounded in the presence of data parallel irregular applications. This paper presents a framework that addresses development and execution of data parallel irregular applications in heterogeneous systems. A unified task-based programming and execution model is proposed, together with inter and intra-device scheduling, which, coupled with a data management system, aim to achieve performance scalability across multiple devices, while maintaining high programming productivity. Intra-device scheduling on wide SIMD/SIMT architectures resorts to consumer-producer kernels, which, by allowing dynamic generation and rescheduling of new work units, enable balancing irregular workloads and increase resource utilization. Results show that regular and irregular applications scale well with the number of devices, while requiring minimal programming effort. Consumer-producer kernels are able to sustain significant performance gains as long as the workload per basic work unit is enough to compensate overheads associated with intra-device scheduling. This not being the case, consumer kernels can still be used for the irregular application. Comparisons with an alternative framework, StarPU, which targets regular workloads, consistently demonstrate significant speedups. This is, to the best of our knowledge, the first published integrated approach that successfully handles irregular workloads over heterogeneous systems.

Abstract
This communication presents an evolutionary software prototype of a user-centered Highly Efficient Pipelined Framework, HEP-Frame, to aid the development of sustainable parallel scientific code with a flexible pipeline structure. HEP-Frame is the result of a tight collaboration between computational scientists and software engineers: it aims to improve scientists coding productivity, ensuring an efficient parallel execution on a wide set of multicore systems, with both HPC and HTC techniques. Current prototype complies with the requirements of an actual scientific code, includes desirable sustainability features and supports at compile time additional plugin interfaces for other scientific fields. The porting and development productivity was assessed and preliminary efficiency results are promising.