Accession Number : AD1051101


Title :   Efficient Numeric and Geometric Computations using Heterogeneous Shared Memory Architectures


Descriptive Note : Technical Report,15 Jul 2014,14 Jul 2017


Corporate Author : University of North Carolina - Chapel Hill Chapel Hill United States


Personal Author(s) : Manocha,Dinesh


Full Text : https://apps.dtic.mil/dtic/tr/fulltext/u2/1051101.pdf


Report Date : 04 Oct 2017


Pagination or Media Count : 54


Abstract : The main goals were to develop new set of algorithms for scientific and geometric computing by exploiting the power and performance efficiency of heterogeneous shared memory architectures. These include multi-core CPUs and many-core GPUs. This involves design of new algorithms, development of appropriate software environments, and demonstrating their potential on a few applications. These applications include ray and path tracing for visual and sound rendering and solving the scientific models for fluid simulation. New heterogeneous algorithms will be developed to exploit CPU and GPU cores in parallel.


Descriptors :   computational fluid dynamics , artificial neural networks , collision avoidance , virtual reality , motion planning , information processing , fluid dynamics , navier stokes equations , information systems , computer graphics , differential equations , CENTRAL PROCESSING UNITS , MEMORY DEVICES , MONTE CARLO METHOD , TRACKING , ALGORITHMS


Subject Categories : Computer Hardware


Distribution Statement : APPROVED FOR PUBLIC RELEASE