Dense matrix multiplication involving a symmetric input matrix (SYMM) is implemented in reference distributed-memory codes with the same data distribution as its general analogue (GEMM). We show that, when the symmetric matrix is dominant, such a 2D block-cyclic (2D BC) scheme leads to a lower arithmetic intensity (AI) of SYMM than that of GEMM by a factor of 2. We propose alternative data distributions preserving the memory benefit of SYMM of storing only half of the matrix while achieving up to the same AI as GEMM. We also show that, in the case we can afford the same memory footprint as GEMM, SYMM can achieve a higher AI. We propose a task-based design of SYMM independent of the data distribution. This design allows for scalable A-stationary SYMM with which all discussed data distributions, may they be very irregular, can be easily assessed. We have integrated the resulting code in a reduction dimension algorithm involving a randomized singular value decomposition dominated by SYMM. An experimental study shows a compelling impact on performance.

With the advent of complex modern architectures, the low-level paradigms long considered sufficient to build High Performance Computing (HPC) numerical codes have met their limits. Achieving efficiency, ensuring portability, while preserving programming tractability on such hardware prompted the HPC community to design new, higher level paradigms while relying on runtime systems to maintain performance. However, the common weakness of these projects is to deeply tie applications to specific expert-only runtime system APIs. The OpenMP specification, which aims at providing common parallel programming means for shared-memory platforms, appears as a good candidate to address this issue thanks to the latest task-based constructs introduced in its revision 4.0. The goal of this paper is to assess the effectiveness and limits of this support for designing a high-performance numerical library, ScalFMM, implementing the fast multipole method (FMM) that we have deeply re-designed with respect to the most advanced features provided by OpenMP 4. We show that OpenMP 4 allows for significant performance improvements over previous OpenMP revisions on recent multicore processors and that extensions to the 4.0 standard allow for strongly improving the performance, bridging the gap with the very high performance that was so far reserved to expert-only runtime system APIs.

Summary High performance fast multipole method is crucial for the numerical simulation of many physical problems. In a previous study, we have shown that task‐based fast multipole method provides the flexibility required to process a wide spectrum of particle distributions efficiently on multicore architectures. In this paper, we now show how such an approach can be extended to fully exploit heterogeneous platforms. For that, we design highly tuned graphics processing unit (GPU) versions of the two dominant operators P2P and M2L) as well as a scheduling strategy that dynamically decides which proportion of subsequent tasks is processed on regular CPU cores and on GPU accelerators. We assess our method with the StarPU runtime system for executing the resulting task flow on an Intel X5650 Nehalem multicore processor possibly enhanced with one, two, or three Nvidia Fermi M2070 or M2090 GPUs (Santa Clara, CA, USA). A detailed experimental study on two 30 million particle distributions (a cube and an ellipsoid) shows that the resulting software consistently achieves high performance across architectures. Copyright © 2015 John Wiley & Sons, Ltd.

SUMMARY We designed a fast parallel simulator that solves the acoustic wave equation on a graphics processing unit (GPU) cluster. Solving the acoustic wave equation in an oil exploration industrial context aims at speeding up seismic modeling and reverse time migration (RTM). We considered a finite difference approach on a regular mesh, in both two‐dimensional and three‐dimensional cases. The acoustic wave equation is solved in a constant density or a variable density domain. All the computations were carried out in single precision (both in the CPU reference implementation and in the GPU implementation), because double precision was not required in our context. We used Compute Unified Device Architecture to take advantage of the GPU computational power. We studied different implementations and their impact on the application performance. The described application handles all the steps of seismic modeling and RTM and is used to solve real‐world problems in an industrial production context. We obtained a speedup of 16 for RTM and up to 43 for the modeling application over a sequential code running on general‐purpose CPUs. A CPU rack versus a GPU rack comparison was described and showed a 4.3 speedup. Copyright © 2011 John Wiley & Sons, Ltd.

We present in this paper multi-thread and multi-process parallelizations of the fast multipole method (FMM) for Laplace equation, for uniform and non uniform distributions. These parallelizations apply to the original FMM formulation and to our new matrix formulation with BLAS (basic linear algebra subprograms) routines. Differences between the multi-thread and the multi-process versions are detailed, and a hybrid MPI-thread approach enables to gain parallel efficiency and memory scalability over the pure MPI one on clusters of SMP nodes. On 128 processors, we obtain 85% (respectively 75%) parallel efficiency for uniform (respectively non uniform) distributions with up to 100 million particles.

In this paper, we present an abstract model to steer legacy coupled simulations that follow the Multiple-SPMD paradigm or the client/server paradigm. This model extends our previous work for the steering of classical SPMD simulations. It describes the application in terms of execution flow, complex distributed data objects and user interactions. Thanks to this abstraction, we define a coordination algorithm that allows us to efficiently interact with the simulation and to overcome the time-coherence problem raised by coupled simulations.

In the context of scientific computing, the computational steering consists in the coupling of numerical simulations with 3D visualization systems through the network. This allows scientists to monitor online the intermediate results of their computations in a more interactive way than the batch mode, and allows them to modify the simulation parameters on-the-fly. While most of existing computational steering environments support parallel simulations, they are often limited to sequential visualization systems. This may lead to an important bottleneck and increased rendering time. To achieve the required performance for online visualization, we have designed the EPSN framework, a computational steering environment that enables to interconnect legacy parallel simulations with parallel visualization systems. For this, we have introduced a redistribution algorithm for unstructured data, that is well adapted to the context of M times N computational steering. Then, we focus on the design of our parallel viewer and present some experimental results obtained with a particle-based simulation in astrophysics