Evaluation of the 3-D finite difference implementation of the acoustic diffusion equation model on massively parallel architectures

The diffusion equation model is a popular tool in room acoustics modeling. The 3-D Finite Difference (3D-FD) implementation predicts the energy decay function and the sound pressure level in closed environments. This simulation is computationally expensive, as it depends on the resolution used to model the room. With such high computational requirements, a high-level programming language (e.g., Matlab) cannot deal with real life scenario simulations. Thus, it becomes mandatory to use our computational resources more efficiently. Manycore architectures, such as NVIDIA GPUs or Intel Xeon Phi offer new opportunities to enhance scientific computations, increasing the performance per watt, but shifting to a different programming model. This paper shows the roadmap to use massively parallel architectures in a 3D-FD simulation. We evaluate the latest generation of NVIDIA and Intel architectures. Our experimental results reveal that NVIDIA architectures outperform by a wide margin the Intel Xeon Phi co-processor while dissipating approximately 50W less (25%) for large-scale input problems.

[1]  Murray Hodgson,et al.  On the use of a diffusion equation for room-acoustic prediction , 2006 .

[2]  Ferenc Molnar,et al.  Simulation of reaction–diffusion processes in three dimensions using CUDA , 2011 .

[3]  Rohit Chandra,et al.  Parallel programming in openMP , 2000 .

[4]  Yao Zhang,et al.  Parallel Computing Experiences with CUDA , 2008, IEEE Micro.

[5]  George Ho,et al.  PAPI: A Portable Interface to Hardware Performance Counters , 1999 .

[6]  Jose J. Lopez,et al.  Some comments about graphic processing unit architectures applied to finite-difference time-domain room acoustics simulation: Present and future trends , 2013 .

[7]  Christian Lengauer,et al.  Optimization of two Jacobi Smoother Kernels by Domain-Specific Program Transformation , 2014 .

[8]  Anas Sakout,et al.  Introducing atmospheric attenuation within a diffusion model for room-acoustic predictions. , 2008, The Journal of the Acoustical Society of America.

[9]  Lasse Natvig,et al.  Temperature effects on on-chip energy measurements , 2013, 2013 International Green Computing Conference Proceedings.

[10]  Lauri Savioja,et al.  Modeling Techniques for Virtual Acoustics , 1999 .

[11]  Ning Xiang,et al.  On boundary conditions for the diffusion equation in room-acoustic prediction: Theory, simulations, and experiments. , 2008, The Journal of the Acoustical Society of America.

[12]  José L. Abellán,et al.  Stencil computations on heterogeneous platforms for the Jacobi method: GPUs versus Cell BE , 2012, The Journal of Supercomputing.

[13]  Apan Qasem,et al.  Understanding stencil code performance on multicore architectures , 2011, CF '11.

[14]  Jose J. Lopez,et al.  A Theoretical Approach to Room Acoustic Simulations Based on a Radiative Transfer Model , 2010 .

[15]  Kevin Skadron,et al.  A performance study of general-purpose applications on graphics processors using CUDA , 2008, J. Parallel Distributed Comput..

[16]  Rezaur Rahman.  Xeon Phi System Software , 2013 .

[17]  Paulius Micikevicius,et al.  3D finite difference computation on GPUs using CUDA , 2009, GPGPU-2.

[18]  Dinesh Manocha,et al.  Use of GPUs in room acoustic modeling and auralization , 2010 .

[19]  Jose J. Lopez,et al.  Implementation and evaluation of a diffusion equation model based on finite difference schemes for sound field prediction in rooms , 2012 .

[20]  P. Morse,et al.  Methods of theoretical physics , 1955 .

[21]  S. Frankel,et al.  Stability conditions in the numerical treatment of parabolic differential equations , 1953 .

[22]  Steven J. Burian,et al.  Assessment of GPU computational enhancement to a 2D flood model , 2011, Environ. Model. Softw..

[23]  Antônio Carlos de Abreu Mól,et al.  GPU-based Monte Carlo simulation in neutron transport and finite differences heat equation evaluation , 2011 .