Optimizing 3D Convolutions for Wavelet Transforms on CPUs with SSE Units and GPUs

Videau, Brice; Marangozova-Martin, Vania; Genovese, Luigi; Deutsch, Thierry

doi:10.1007/978-3-642-40047-6_82

Cited by 5 publications

(3 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…GPU acceleration has been achieved for wave function based DFT methods, e.g. plane wave methods, wavelets, grid and local orbitals [11,20,22,24]. Closer to this work are projects in the class of linear-scaling methods like SIESTA or CP2K [6,21].…”

Section: Related Workmentioning

confidence: 98%

Addressing Materials Science Challenges Using GPU-accelerated POWER8 Nodes

Baumeister

Bornemann

Bühler

et al. 2016

Euro-Par 2016: Parallel Processing

View full text Add to dashboard Cite

Materials research is an area that is expected to strongly benefit from the growing performance capabilities of future supercomputers towards exascale. Density functional theory (DFT) has become one of the most important methods for numerical materials science. In this paper we present results of a performance model based analysis of a particular, scalable DFT-based application on GPU-accelerated compute nodes with POWER8 processors. These technologies are part of a future roadmap for pre-exascale architectures. With power consumption becoming a major design constraint, we also determine the energy required for executing the most performance critical kernel.

show abstract

Section: Related Workmentioning

confidence: 98%

Addressing Materials Science Challenges Using GPU-accelerated POWER8 Nodes

Baumeister

Bornemann

Bühler

et al. 2016

Euro-Par 2016: Parallel Processing

View full text Add to dashboard Cite

show abstract

“…The CPU convolutions of BigDFT have thus been thoroughly optimized. In a recent paper (Videau et al, 2013), the optimization of the CPU convolutions of BigDFT has been extensively considered. One example of a specific convolution, called MagicFilter (Genovese et al, 2010), can be seen in Listing 16.…”

Section: Creating An Autotuned Convolution Library For Bigdft Using Bmentioning

confidence: 99%

Boast

Videau

Pouget

Genovese

et al. 2017

The International Journal of High Performance Computing Applica

Self Cite

View full text Add to dashboard Cite

The portability of real high-performance computing (HPC) applications on new platforms is an open and very delicate problem. Especially, the performance portability of the underlying computing kernels is problematic as they need to be tuned for each and every platform the application encounters. This article presents BOAST, a metaprogramming framework dedicated to computing kernels. BOAST allows the description of a kernel and its possible optimizations using a domain-specific language. BOAST runtime will then compare the different versions'performance as well as verify their exactness. BOAST is applied to three use cases: a Laplace kernel in OpenCL and two HPC applications BigDFT (electronic density computation) and SPECFEM3D (seismic and wave propagation).

show abstract

“…Several works have studied data layout restructuring for specific applications [21,12] and for stencils [9]. In a recent work [1] the authors proposed a framework to analyze binary codes, and to formulate user-targeted hints about SIMDization potentials and hindrances.…”

Section: Introductionmentioning

confidence: 99%

Rewriting System for Profile-Guided Data Layout Transformations on Binaries

Haine

Aumage²,

Barthou

2017

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Careful data layout design is crucial for achieving high performance. However exploring data layouts is time-consuming and errorprone, and assessing the impact of a layout transformation on performance is difficult without performing it. We propose to guide application programmers through data layout restructuring by providing a comprehensive multidimensional description of the initial layout, built from trace analysis, and then by giving a performance evaluation of the transformations tested and an expression of each transformed layout. The programmer can limit the exploration to layouts matching some patterns. We apply this method to two multithreaded applications. The performance prediction of multiple transformations matches within 5% the performance of hand-transformed layout code.

show abstract

Optimizing 3D Convolutions for Wavelet Transforms on CPUs with SSE Units and GPUs

Cited by 5 publications

References 15 publications

Addressing Materials Science Challenges Using GPU-accelerated POWER8 Nodes

Addressing Materials Science Challenges Using GPU-accelerated POWER8 Nodes

Boast

Rewriting System for Profile-Guided Data Layout Transformations on Binaries

Contact Info

Product

Resources

About