Due to its architecture,the graphics processing unit(GPU)is specially well-suited to address problems that can be expressed as data-parallel computations with high arithmetic intensity.One example of such problem is t...Due to its architecture,the graphics processing unit(GPU)is specially well-suited to address problems that can be expressed as data-parallel computations with high arithmetic intensity.One example of such problem is the Boundary Elements Method(BEM).This work addresses the implementation of the direct version of BEM for 2D elastostatics.For the present implementation,constant boundary elements are used.According to the formulation of BEM,every term of both influence matrices(G^ij and H^ij)is independent of each other.In classical CPU serial implementations,these terms are calculated in a sequence of two loops:for the field point i and for the source point j.On the other hand,from the point of view of the GPU parallel processing paradigm,the calculation of every one of these terms can be assigned to a thread(GPU's elementary unit of calculation)and calculated simultaneously.The transposition of the influence equation to an algebraic linear system of equations is also parallelized.Standard Gaussian quadrature is applied to integrate each term of influence matrices.The code was developed on a NVidia CUDA programming environment and executed on a GeForce GTX 280 graphics card hosted by a regular Intel Core2Duo CPU.The efficiency of the implemented strategies is investigated by solving a classical elastostatics problem.展开更多
Over the forty-year history of interactive computer graphics, there have beencontinuous advances, but at some stage this progression must terminate with images beingsufficiently realistic for all practical purposes. H...Over the forty-year history of interactive computer graphics, there have beencontinuous advances, but at some stage this progression must terminate with images beingsufficiently realistic for all practical purposes. How much detail do we really need? Polygon countsover a few million imply that on average each polygon paints less than a single pixel, making useof polygon shading hardware wasteful. We consider the problem of determining how much realism isrequired for a variety of applications. We discuss how current trends in computer graphics hardware,and in particular of graphics cards targeted at the computer games industry, will help or hinderachievement of these requirements. With images now being so convincingly realistic in many cases,critical faculties are often suspended and the images are accepted as correct and truthful althoughthey may well be incorrect and sometimes misleading or untruthful. Display resolution has remainedlargely constant in spatial terms for the last twenty years and in terms of the number of pixels hasincreased by less than an order of magnitude. If the long-promised breakthroughs in displaytechnology are finally realised, how should we use the increased resolution?展开更多
One of the problems in virtual globes technologies is the real-time representation of vegetal species.In forest or garden representations,the low detailed plants produce a lack of realism.Efficient techniques are requ...One of the problems in virtual globes technologies is the real-time representation of vegetal species.In forest or garden representations,the low detailed plants produce a lack of realism.Efficient techniques are required to achieve accurate interactive visualisation due to the great number of polygons the vegetal species have.This article presents a multi-resolution model based on a geometric representation of vegetal species that allows the application to perform the progressive transmission of the model,that is,the transmission of a simple representation followed by successive refinements of it.It has a hardware-oriented design in order to obtain interactive frame rates.The geometric data of the objects are stored in the graphics processing unit and,moreover,the change from one approximation to another is obtained by performing mathematical calcula-tions in this graphics hardware.The multi-resolution model presented here enables instancing:as many vegetal species as desired can be rendered with different levels of detail,while all of them are accessing the same geometric data.This model has been used to build a real-time representation of a not imaginary scenario.展开更多
With increasing demands of virtual reality(VR) applications, efficient VR rendering techniques are becoming essential. Because VR stereo rendering has increased computational costs to separately render views for the l...With increasing demands of virtual reality(VR) applications, efficient VR rendering techniques are becoming essential. Because VR stereo rendering has increased computational costs to separately render views for the left and right eyes, to reduce the rendering cost in VR applications, we present a novel traversal order for tile-based mobile GPU architectures: Z^2 traversal order. In tile-based mobile GPU architectures,a tile traversal order that maximizes spatial locality can increase GPU cache efficiency. For VR applications, our approach improves upon the traditional Z order curve.We render corresponding screen tiles in left and right views in turn, or simultaneously, and as a result, we can exploit spatial adjacency of the two tiles. To evaluate our approach, we conducted a trace-driven hardware simulation using Mesa and a hardware simulator. Our experimental results show that Z^2 traversal order can reduce external memory bandwidth requirements and increase rendering performance.展开更多
文摘Due to its architecture,the graphics processing unit(GPU)is specially well-suited to address problems that can be expressed as data-parallel computations with high arithmetic intensity.One example of such problem is the Boundary Elements Method(BEM).This work addresses the implementation of the direct version of BEM for 2D elastostatics.For the present implementation,constant boundary elements are used.According to the formulation of BEM,every term of both influence matrices(G^ij and H^ij)is independent of each other.In classical CPU serial implementations,these terms are calculated in a sequence of two loops:for the field point i and for the source point j.On the other hand,from the point of view of the GPU parallel processing paradigm,the calculation of every one of these terms can be assigned to a thread(GPU's elementary unit of calculation)and calculated simultaneously.The transposition of the influence equation to an algebraic linear system of equations is also parallelized.Standard Gaussian quadrature is applied to integrate each term of influence matrices.The code was developed on a NVidia CUDA programming environment and executed on a GeForce GTX 280 graphics card hosted by a regular Intel Core2Duo CPU.The efficiency of the implemented strategies is investigated by solving a classical elastostatics problem.
文摘Over the forty-year history of interactive computer graphics, there have beencontinuous advances, but at some stage this progression must terminate with images beingsufficiently realistic for all practical purposes. How much detail do we really need? Polygon countsover a few million imply that on average each polygon paints less than a single pixel, making useof polygon shading hardware wasteful. We consider the problem of determining how much realism isrequired for a variety of applications. We discuss how current trends in computer graphics hardware,and in particular of graphics cards targeted at the computer games industry, will help or hinderachievement of these requirements. With images now being so convincingly realistic in many cases,critical faculties are often suspended and the images are accepted as correct and truthful althoughthey may well be incorrect and sometimes misleading or untruthful. Display resolution has remainedlargely constant in spatial terms for the last twenty years and in terms of the number of pixels hasincreased by less than an order of magnitude. If the long-promised breakthroughs in displaytechnology are finally realised, how should we use the increased resolution?
基金This work was supported by the Spanish Ministry of Science and Technology(Project TIN2010-21089-C03-03)Feder Funds and Generalitat Valenciana(Project PROMETEO/2010/028).
文摘One of the problems in virtual globes technologies is the real-time representation of vegetal species.In forest or garden representations,the low detailed plants produce a lack of realism.Efficient techniques are required to achieve accurate interactive visualisation due to the great number of polygons the vegetal species have.This article presents a multi-resolution model based on a geometric representation of vegetal species that allows the application to perform the progressive transmission of the model,that is,the transmission of a simple representation followed by successive refinements of it.It has a hardware-oriented design in order to obtain interactive frame rates.The geometric data of the objects are stored in the graphics processing unit and,moreover,the change from one approximation to another is obtained by performing mathematical calcula-tions in this graphics hardware.The multi-resolution model presented here enables instancing:as many vegetal species as desired can be rendered with different levels of detail,while all of them are accessing the same geometric data.This model has been used to build a real-time representation of a not imaginary scenario.
文摘With increasing demands of virtual reality(VR) applications, efficient VR rendering techniques are becoming essential. Because VR stereo rendering has increased computational costs to separately render views for the left and right eyes, to reduce the rendering cost in VR applications, we present a novel traversal order for tile-based mobile GPU architectures: Z^2 traversal order. In tile-based mobile GPU architectures,a tile traversal order that maximizes spatial locality can increase GPU cache efficiency. For VR applications, our approach improves upon the traditional Z order curve.We render corresponding screen tiles in left and right views in turn, or simultaneously, and as a result, we can exploit spatial adjacency of the two tiles. To evaluate our approach, we conducted a trace-driven hardware simulation using Mesa and a hardware simulator. Our experimental results show that Z^2 traversal order can reduce external memory bandwidth requirements and increase rendering performance.
