“Working in close collaboration with Intel Labs Parallel Computing Lab, performing a series of architecture-aware optimizations, the team was able to scale the complexity of science and sustained performance to an unprecedented level. SeisSol sustained 8.6 PFLOPS (double precision), almost equivalent 8.6 quadrillion calculations per second when processing seismic wave phenomena using half of the Tianhe-2 supercomputer.
“This paper provides information and benchmarks necessary to make the choice of the best file system for a given application from a number of the available options: RAM disks, virtualized local hard drives, and distributed storage shared with NFS or Lustre. We report benchmarks of I/O performance and parallel scalability on Intel Xeon Phi coprocessors, strengths and limitations of each option.”
Over at Typhoon Computing, Michel Müller writes programmers looking to port their code to accelerators now have a new tool called Hybrid Fortran. “This python-based preprocessor parses annotations together with your Fortran code structure, declarations, accessors and procedure calls, and then writes separate versions of your code – once for CPU with OpenMP parallelization and once for GPU with CUDA Fortran.”
In this video from ISC’14, Alex Heinecke from Intel and Sebastian Rettenberger from the Technical University of Munich describe their award-winning paper on volcano simulation. “Seismic simulations in realistic 3D Earth models require peta- or even exascale compute power to capture small-scale features of high relevance for scientific and industrial applications. In this paper, we present optimizations of SeisSol — a seismic wave propagation solver based on the Arbitrary high-order accurate DERivative (ADER) Discontinuous Galerkin method on fully adaptive, unstructured tetrahedral meshes — to run simulations under production conditions at petascale performance.”