300x Matlab
نویسنده
چکیده
The true costs of high performance computing are currently dominated by software. Addressing these costs requires shifting to high productivity languages such as Matlab. MatlabMPI is a Matlab implementation of the Message Passing Interface (MPI) standard and allows any Matlab program to exploit multiple processors. The performance has been tested on both shared and distributed memory parallel computers (Sun, SGI, HP, IBM and Linux). A test image filtering application using MatlabMPI achieved a speedup of 300 using 304 CPUs and 15% of the theoretical peak (450 Gigaflops) on an IBM SP2 at the Maui High Performance Computing Center. In addition, this entire parallel benchmark application was implemented in 70 software-lines-of-code (SLOC) yielding 0.85 Gigaflop/SLOC or 4.4 CPUs/SLOC. The MatlabMPI software will be available for download at hpcmo.hpc.mil.
منابع مشابه
Analysis of Micromachined Capacitive Incremental Position Sensor
This article presents an analysis for two related concepts of a capacitive incremental position sensor. In Incremental Capacitance Measurement Mode the periodic change in capacitance is measured to determine the relative displacement between two periodic geometries S1 and S2 with gap-distance of ~ 1 μm. In Constant Capacitance Measurement Mode the distance between S1 and S2 is controlled to k...
متن کاملFast 3D Object Recognition of Rotationally Symmetric Objects
In this paper we extend a recent approach for 3D object recognition in order to deal with rotationally symmetric objects, which are frequent in daily environments. We base our work in a recent method that represents objects using a hash table of shape features, which in the case of symmetric objects contains redundant information. We propose a way to remove redundant features by adding a weight...
متن کاملRealtime Unsupervised Texture Segmentation Using Graphics Hardware
General purpose computation on graphics processing units (GPGPU) has opened up a host of possibilities for high performance computing on commodity hardware. We show how an interesting texture segmentation algorithm can achieve 35x50x speedups on the GPU. We also show that portions of the algorithm can even approach a 300x speedup. We also demonstrate that portions of the algorithm that form bot...
متن کاملPlanar micro-optic solar concentrator.
We present a new approach to solar concentration where sunlight collected by each lens in a two-dimensional lens array is coupled into a shared, planar waveguide using localized features placed at each lens focus. This geometry yields a thin, flat profile for moderate concentration systems which may be fabricated by low-cost roll manufacture. We provide analyses of tradeoffs and show optimized ...
متن کاملPower-Aware Acoustic Processing
We investigated the tradeoffs between accuracy and battery-energy longevity of acoustic beamforming on disposable sensor nodes subject to varying key parameters: 1) number of microphones, 2) duration of sampling, 3) number of search angles, and 4) CPU clock speed. Beyond finding the most energy efficient implementation of the beamforming algorithm at a specified accuracy, we seek to enable appl...
متن کامل