The omission of nonlinear effects in large-scale 3D ground motion estimation, which are particularly challenging due to memory and scalability issues, can result in costly misguidance for structural design in earthquake-prone regions. We have implemented nonlinearity using a Drucker-Prager yield condition in AWP-ODC and further optimized the CUDA kernels to more efficiently utilize the GPU's memory bandwidth. The application has resulted in a significant increase in the model region and accuracy for state-of-the-art earthquake simulations in a realistic earth structure, which are now able to resolve the wavefield at frequencies relevant for the most vulnerable buildings (> 1 Hz) while maintaining the scalability and efficiency of the method. We successfully run the code on 4,200 Kepler K20X GPUs on NCSA Blue Waters and OLCF Titan to simulate a M 7.7 earthquake on the southern San Andreas fault with a spatial resolution of 25 m for frequencies up to 4 Hz.

We have developed a highly scalable and efficient GPU-based finite-difference code (AWP) for earthquake simulation that implements high throughput, memory locality, communication reduction and communication/computation overlap and achieves linear scalability on Cray XK7 Titan at ORNL and NCSA's Blue Waters system. We simulate realistic 0-10 Hz earthquake ground motions relevant to building engineering design using high-performance AWP. Moreover, we show that AWP provides a speedup by a factor of 110 in key strain tensor calculations critical to probabilistic seismic hazard analysis (PSHA). These performance improvements to critical scientific application software, coupled with improved co-scheduling capabilities of our workflow-managed systems, make a statewide hazard model a goal reachable with existing supercomputers. The performance improvements of GPU-based AWP are expected to save millions of core-hours over the next few years as physics-based seismic hazard analysis is developed using heterogeneous petascale supercomputers.

Petascale simulations are needed to understand the rupture and wave dynamics of the largest earthquakes at shaking frequencies required to engineer safe structures (> 1 Hz). Toward this goal, we have developed a highly scalable, parallel application (AWP-ODC) that has achieved “M8”: a full dynamical simulation of a magnitude-8 earthquake on the southern San Andreas fault up to 2 Hz. M8 was calculated using a uniform mesh of 436 billion 40-m3cubes to represent the three-dimensional crustal structure of Southern California, in a 800 km by 400 km area, home to over 20 million people. This production run producing 360 sec of wave propagation sustained 220 Tflop/s for 24 hours on NCCS Jaguar using 223,074 cores. As the largest-ever earthquake simulation, M8 opens new territory for earthquake science and engineering - the physics-based modeling of the largest seismic hazards with the goal of reducing their potential for loss of life and property.