Computing room acoustics with CUDA - 3D FDTD schemes with boundary losses and viscosity

Abstract

In seeking to model realistic room acoustics, direct numerical simulation can be employed. This paper presents 3D Finite Difference Time Domain schemes that incorporate losses at boundaries and due to the viscosity of air. These models operate within a virtual room designed on a detailed floor plan. The schemes are computed at 44.1kHz, using large-scale data sets containing up to 100 million points each. A performance comparison is made between serial computation in C, and parallel computation using CUDA on GPUs, showing up to 80 times speed-ups. Testing on two different Nvidia Tesla cards shows the benefits of the latest FERMI architecture for double precision floating-point computation.

GPU-Based Dynamic Wave Field Synthesis Using Fractional Delay Filters and Room Compensation

“…Accelerators of this type have already been applied to different problems in acoustics and audio processing. Some applications include room acoustics modeling [24], [25], [26], [27], additive synthesis [28], [29], full 3-D model of drums in a large virtual room [30], sliding phase vocoder [31], beamforming [32], audio rendering [33], [34], [35], multichannel IIR filtering of audio signals [36], dynamic range reduction using multiple allpass filters [37], and adaptive filtering [38], [39], [40].…”

Intel Many-Integrated Core (MIC) architecture-based parallel computation and optimization of finite-difference time-domain (FDTD) schemes for acoustic simulation

“…Fully optimized implementation for the FDTD(2,2), FDTD(2,4), and WE-FDTD(2,2) schemes gain 19.1, 20.3, and 9.3 times speedup against the initial implementation on the Intel MIC, respectively. The performance that was achieved for the FDTD(2,2), FDTD (2,4), and WE-FDTD(2,2) schemes is 45.7, 77.9, and 113 GFlops on the Intel MIC, respectively. Similarly, on the CPU, the gain in speedup was 1.7, 1.6, and 1.7 times, achieving 6.8, 10.1, and 18.4 GFlops, respectively.…”

“…(1) and (2). The FDTD(2,4) scheme is obtained similarly by employing a second-order central difference approximation for the time derivative and a fourth-order central difference approximation for the spatial derivative.…”

“…The movement toward applying GPU to scientific computing has been ongoing since 2005 [1]. A number of studies have reported the application of GPU [2][3][4] to largescale and long-time acoustic simulation for finite-difference time-domain (FDTD) methods [5][6][7]. On the other hand, the first generation of Intel MIC architecture is newer than GPU because it was only released in 2013.…”

GPU Implementation of Multichannel Adaptive Algorithms for Local Active Noise Control

“…A general overview of audio signal processing on the GPU is given in [14] and [15]. Moreover, there are many recent contributions that leverage GPUs to accelerate acoustic and audio simulations or realtime applications like: room acoustics [16], acoustics likelihood computation [17], speech recognition [18], RIR (Room Impulse Response) reshaping [19], beamforming [20], sound localization [21] or wave-field synthesis [22]. Furthermore, the filtering on GPU where real-time filtering of multiple data is carried out concurrently has recently been introduced in [23], [24].…”