An explicit dynamics GPU structural solver for thin shell finite elements

Abstract

With the availability of user oriented software tools, dedicated architectures, such as the parallel computing platform and programming model CUDA (Compute Unified Device Architecture) released by NVIDIA, one of the main producers of graphics cards, and of improved, highly performing GPU (Graphics Processing Unit) boards, GPGPU (General Purpose programming on GPU) is attracting increasing interest in the engineering community, for the development of analysis tools suitable to be used in validation/verification and virtual reality applications. For their inherent explicit and decoupled structure, explicit dynamics finite element formulations appear to be particularly attractive for implementations on hybrid CPU/GPU or pure GPU architectures. The issue of an optimized, double-precision finite element GPU implementation of an explicit dynamics finite element solver for elastic shell problems in small strains and large displacements and rotations, using unstructured meshes, is here addressed. The conceptual difference between a GPU implementation directly adapted from a standard CPU approach and a new optimized formulation, specifically conceived for GPUs, is discussed and comparatively assessed. It is shown that a speedup factor of about 5 can be achieved by an optimized algorithm reformulation and careful memory management. A speedup of more than 40 is achieved with respect of state-of-the art commercial codes running on CPU, obtaining real-time simulations in some cases, on commodity hardware. When a last generation GPU board is used, it is shown that a problem with more than 16 millions degrees of freedom can be solved in just few hours of computing time, opening the way to virtualization approaches for real large scale engineering problems.

Rupture risk in abdominal aortic aneurysms: A realistic assessment of the explicit GPU approach

“…With the release of Compute Capability 1.3 devices in late 2008, fp64 computation is supported, but is still largely absent in literature. Bartezzaghi et al (2015) first addressed this issue, and with the exception of Strbac et al (2015) and Strbac et al (2016), the authors are unaware of reports using doubleprecision computation in our context. It is well-known that double-precision computation, particularly in atomic operations, is significantly slower but mandatory in accurate and general engineering application.…”

“…Reports of large (up to two orders of magnitude) speed-ups can be found in the literature for the TLED method Taylor et al (2009), Strbac et al (2015), Strbac et al (2016), Bartezzaghi et al (2015). Such claims are often based on simplifications, some introduced in the following paragraphs.…”

“…Further, benchmarking reports on GPU performance use simple or ideal (Bartezzaghi et al, 2015;Comas et al, 2008;Strbac et al, 2015) geometries. Meshing of complex geometries often requires misshaped elements (that limit usable step sizes in explicit analysis), also detailed further in Section 2.3.1.…”

Efficient topology optimization using GPU computing with multilevel granularity

“…These graphics devices are also successfully used for real-time simulation and haptic feedback of soft tissue deformations [27,28], which are especially useful for the development of realistic simulators. Besides, relevant results are obtained for the structural solver of finite element explicit dynamics problems using GPU computing [29]. These graphics cards have also shown promising results in heterogeneous systems addressing large-scale problems in Finite Element Analysis (FEA) [30].…”

An approach to enhance the performance of large‐scale structural analysis on CPU‐MIC heterogeneous clusters

“…Due to the limitation of computing power and memory space, the traditional computer cannot meet the requirements of large‐scale structural analyses . As a result, supercomputers with many processors and memories are employed …”