High-Performance Computation of Bézier Surfaces on Parallel and Heterogeneous Platforms

Abstract

Bézier surfaces are mathematical tools employed in a wide variety of applications. Some works in the literature propose parallelization strategies to improve performance for the computation of Bézier surfaces. These approaches, however, are mainly focused on graphics applications and often are not directly applicable to other domains. In this work, we propose a new method for the computation of Bézier surfaces, together with approaches to efficiently map the method onto different platforms (CPUs, discrete and integrated GPUs). Additionally, we explore CPU-GPU cooperation mechanisms for computing Bézier surfaces using two integrated heterogeneous systems with different characteristics. An exhaustive performance evaluation-including different data-types, rendering and several hardware platforms-is performed. The results show that our method achieves speedups as high as 3.12x (double-precision) and 2.47x (single-precision) on CPU, and 3.69x (double-precision) and 13.14x (single-precision) on GPU compared to other methods in the literature. In heterogeneous platforms, the CPU-GPU cooperation increases the performance up to 2.09x with respect to the GPU-only version. Our method and the 123Int J Parallel Prog associated parallelization approaches can be easily employed in domains other than computer-graphics (e.g., image registration, bio-mechanical modeling and flow simulation), and extended to other Bézier formulations and Bézier constructions of higher order than surfaces.

Modeling Variation in Multi-Station Compliant Assembly Using Parametric Space Envelope

“…For the developed method, the bulk calculation involves the cubic B-Spline computation. Many techniques and architectures have been developed in the past decade to improve this computation efficiency [50][51][52][53]. Now it is possible to achieve real-time (or near real-time) fast implementation [54].…”

Optimizing a medical image registration algorithm based on profiling data for real-time performance

“…From several works [31,32,26], one can realize that in the field of medical image processing and analysis, it is not common for the developers to use tools that detect computationally costly functions in their algorithms. However, several studies have been conducted to address performance issues in image registration algorithms using high-performance computing [33,13,15]. For example, Shackleford et al [14] presented a comprehensive survey of non-rigid registration algorithms that are suitable for use in modern multi-core architectures.…”

“…Hence, this task is well-known as one of the most time-consuming tasks that can be found in medical image analysis [11,12]. However, with the development of multi-core processor architecture, several solutions have been proposed that realize non-rigid image registration algorithms on multi-core CPUs [13,14,15]. Multi-core architecture mainly aim at improving the performance of highly demanded applications by exploiting parallelism.…”