统一渲染架构GPU图形处理量化性能模型研究

统一渲染架构GPU为图形处理提供了丰富的运算、存储资源,也对软件优化提出了更高要求。为了有效地进行性能设计和优化,针对统一渲染架构实现的GPU提出一种量化的图形处理性能模型,在深入研究统一渲染架构GPU架构和工作原理基础上,分析影响图形处理的各种因素:图形指令生成、主机接口数据传输、图形指令解析、图形处理流水数据吞吐和统一染色阵列处理能力。通过仿真验证表明,在研制自主知识产权GPU过程中,采用本方法设计各部分性能指标,评估统一染色GPU图形处理性能与实测相比,误差小于7.5%。     

基于GPU的遥感影像数据融合IHS变换算法

提出基于图形处理单元（GPU）的遥感影像IHS融合算法,利用图形硬件的可编程渲染器和其处理数据的并行性,把IHS的正反变换映射到GPU中进行计算。应用RTT和MRT技术实现IHS正反变换中3个分量的并行渲染输出,加速计算过程。实验结果表明,在数据量较大时,该算法的处理速度比基于CPU的算法速度更快。     

基于DCT并行加速算法图像渲染平台系统设计

此次主要研究了基于GPU的集群渲染系统平台设计;为了提高平台的工作效率、增强集群渲染系统平台的数据传输能力,提出了一种采用DCT变换的方法来加速图像渲染速度;该方法利用DCT变换算法加速图像的实时压缩,加入CPU监控器和任务分配器模块,让GPU和CPU共同承担了绘图和渲染的目的,这样有效地降低处理流程对CPU的占用,实现了三维绘图和特效渲染的加速;为了验证平台的有效性以及图像压缩处理的效果,做了相应的功能验证;对640×480的RCB图像使用上述压缩方法和JPEG标准库在不同压缩设置下进行实验;仿真实验结果表明所提方案具有更高的压缩效率。     

基于GPU的COONS曲面片加速计算

随着可编程图形处理单元(GPU)性能的增强,许多基于GPU的三维几何造型系统的应用也与日俱增。Coons曲面片作为三维建模中重要的基本单元,在农作物的建模仿真以及曲面造型中是重要的。为了适应于GPU的通用计算,对传统Coons曲面片方程进行分解,从而并行构造出Coons曲面片。通过实验表明,随着插值点数的增加,片元程序执行的时间依然保持一定程度的稳定,并没有出现明显的增加;GPU上执行时间的增长速度明显低于在CPU上的增长速度。     

GSim:支持GPU加速软件过程仿真框架

为了提高软件过程仿真的效率,提出了一种基于图形处理单元(graphic processing unit,GPU)加速的仿真框架.该框架利用图形化语言和随机参数来描述一个过程模型,将模型转换为RansomSpec字节码从而可以在GPU平台上运行,以期借助GPU平台的高并行特性提高原有仿真算法的效率.实验结果表明,通过这种框架,基于GPU的随机软件过程仿真与传统基于CPU串行的仿真算法相比在效率上提高一个数量级.     

基于混合模式匹配算法的网络入侵检测

为了提升中央处理单元(CPU)和图形处理单元(GPU)协同检测网络入侵的性能,本文提出了一种具有数据包有效载荷长度约束的CPU/GPU混合模式匹配算法(LHPMA)。在分析CPU/GPU混合模式匹配算法(HPMA)的基础上,设计了长度约束分离算法(LBSA)对传入数据包进行提前分类。利用CPU中的预过滤缓冲区对较长数据包进行快速预过滤,结合全匹配缓冲区将较短数据包直接分配给GPU进行全模式匹配,通过减少有效载荷长度的多样性,提升了CPU/GPU协同检测网络入侵的性能。实验结果表明,LHPMA增强了HPMA的处理性能,充分发挥了GPU并行处理较短数据包的优势,并且LHPMA提高了网络入侵检测的吞吐量。     

负载平衡在三维渲染中的应用

该文介绍一种基于调整CPU和GPU(GraphicsProcessUnit)数据处理量来达到负载平衡,进而加速三维场景渲染的方法。现代GPU芯片制造技术发展非常迅速,直接导致GPU性能迅速增长,传统的通过简单地减少显卡工作量来优化渲染速度的思想已经不是十分合适了,取而代之的应该是平衡CPU和GPU的数据处理量来优化渲染,基于这个思想我们展开了一点讨论并对传统的算法作出一些修改,通过实验比较了CPU与GPU数据处理量平衡和非平衡时的渲染速度,实验结果证明了基于平衡数据处理量来进行优化渲染速度的方法是切实可行的。     

基于GPU的碳纳米管分子动力学并行仿真

基于计算机的分子动力学仿真具有理论分析方法和实验方法无法比拟的优点,但分子动力学仿真算法计算量非常大,特别是在对碳纳米管的大规模粒子数进行仿真处理时,普通的基于CPU的串行算法执行效率低且耗时多。为此,提出基于统一计算设备架构的碳纳米管分子动力学的图形处理单元( GPU)并行算法,设计并实现仿真算法中适合GPU并行运算的分裂算法,将具有竞争资源的运算以非竞争方式运行。实验结果表明,与CPU串行仿真算法相比,分裂算法的运算速度较快,且在只有16个GPU流处理器显卡上可获得十多倍的加速比。     

基于OpenCL的GPU加速三维时域有限差分电磁场仿真算法研究

提出了一种基于开放运算语言（OpenCL）的GPU加速三维时域有限差分（FDTD）电磁场仿真计算的方法．该方法利用图形处理单元（GPU）的并行处理特性并结合OpenCL接口标准实现了时域卷积完全匹配层（CPML）吸收边界条件的三维FDTD的高性能加速计算．首先设置FDTD仿真参数并动态申请内存空间,然后初始化OpenCL的计算参数,对三维电磁模型基于OpenCL进行FDTD加速仿真．本方法显著提升了FDTD电磁场仿真速度,与利用CPU计算相比速度提升可达5-8倍,且具有CPML吸收边界条件,可以模拟电磁波在自由空间的传播;基于OpenCL编译的语言程序可以运行在CPU或GPU硬件上,并可充分发挥多核CPU的并行计算能力,使得FDTD电磁场仿真具有更广泛的实际应用．     

GPU构网的GeoMipMap地形无缝绘制算法

针对传统CPU实时构网算法和预处理阶段构网算法速度较慢问题,提出一种GPU构网的GeoMipMap地形渲染算法.算法的构网阶段由GPU实现,将CPU从繁重的构网工作中解放出来,并大幅度减少CPU向GPU传输的数据量,提高地形的渲染速度.整个地形分成大小相等的若干地形块,每个地形块又分为内部及四条边共五部分,对这五部分按分辨率不同分成多个细节层次,为每个细节层次计算空间误差,渲染时各部分根据屏幕投影误差选择细节层次,所构网格更加符合地形表面特征.考虑到GPU构网算法的高度并行性,采用一种新的裂缝处理方式,四条边的屏幕投影误差以边上顶点的空间误差计算,使得相邻块对于共享边的细节层次的计算结果相同,从而保证相邻块间无裂缝,且网格连续.实验结果表明该方法能够以较高的质量完成大规模地形的实时平滑漫游.     

并行化水体表面动态多分辨率物理仿真模型

大面积水体仿真中的细节捕捉较为困难,且渲染效率容易受限。为此,提出一种基于图形处理单元(GPU)的水体表面动态多分辨率物理仿真模型。采用细节层次渐进网格划分方法对水体表面进行建模,利用统一计算构架和GPU技术加速纳维-斯托克斯方程的求解。实验结果表明,随着初始网格分辨率的提高,该模型的加速比逐渐加大,在分辨率达到1 024×1 024的情况下,能保证30帧左右的渲染 效率。     

基于CUDA的弱可压SPH流体建模与仿真

为了实现小尺度范围流体场景的实时、真实感模拟,采用弱可压SPH方法对水体进行建模,提出了流体计算的CPU GPU混合架构计算方法。针对邻域粒子查找算法影响流体计算效率的问题,采用三维空间网格对整个模拟区域进行均匀网格划分,利用并行前缀求和和并行计数排序实现邻域粒子的查找。最后,采用基于CUDA并行加速的Marching Cubes算法实现流体表面提取,利用环境贴图表现流体的反射和折射效果,实现流体表面着色。实验结果表明,所提出的流体建模和模拟算法能实现小尺度范围流体的实时计算和渲染,绘制出水的波动、翻卷和木块在水中晃动的动态效果,当粒子数达到1 048 576个时,GPU并行计算方法相较CPU方法的加速比为60.7。     

混合插值法重构近地表模型

当控制点多和网格稠密时,基于薄板样条(TPS)插值的近地表模型重构往往很耗时,影响了静校正中近地表建模的效率.针对此问题,采用一种TPS插值和三次样条插值相结合的混合插值法重构近地表模型.首先利用矩阵递归LU分解及GPU加速的LU分解算法求解大型线性方程组,建立TPS插值函数;然后在X和Y方向上使用适当的步长对网格进行抽稀,运用TPS插值函数计算稀疏网格点的值,再通过稀疏网格点建立三次样条插值函数并计算剩余网格点的值;最后用OpenGL实现近地表模型的三维可视化.实验结果表明,文中算法提高了近地表模型重构的速度,其精度接近TPS插值精度.     

图形硬件加速的实时水面绘制

实时生成具有真实感效果的水面是计算机图形学中的研究热点和难点之一。文章介绍了一个利用可编程图形硬件来实现水面实时生成和绘制的系统,绘制过程主要分两个方面:水面的建模和水面光照效果的实现。通过基于空间域的快速傅立叶变换技术来实现水面的建模,通过凹凸纹理贴图和投影纹理技术来实现水面的反射、折射和菲涅耳等水面光照效果。绘制过程主要在图形处理器中实现,从而保证了算法的实时性。在现有的PC机和可编程图形硬件加速卡上能达到每秒30帧以上的绘制速度。     

实时水面模拟方法研究

提出了一种基于GPU的水面实时模拟方法。该方法不依赖于噪声图,而实现了实时的水波生成、折射和反射效果的菲涅耳合成以及水面光照模型的计算。利用GPU在片段处理前的光栅化处理,该方法渲染负荷不会因水面大小和精度而增大。且依赖GPU的高速计算能力,方法可以达到实时。     

Particle transport with unstructured grid on GPU

The method of discontinuous finite element discrete ordinates which involves inverting an operator by iteratively sweeping across a mesh from multiple directions is commonly used to solve the time-dependent particle transport equation. Graphics Processing Unit (GPU) provides great faculty in solving scientific applications. The particle transport with unstructured grid bringing forward several challenges while implemented on GPU. This paper presents an efficient implementation of particle transport with unstructured grid under 2D cylindrical Lagrange coordinates system on a fine-grained data level parallelism GPU platform from three aspects. The first one is determining the sweep order of elements from different angular directions. The second one is mapping the sweep calculation onto the GPU thread execution model. The last one is efficiently using the on-chip memory to improve performance. As to the authors? knowledge, this is the first implementation of a general purpose particle transport simulation with unstructured grid on GPU. Experimental results show that the performance speedup of NVIDIA M2050 GPU with double precision floating operations ranges from 11.03 to 17.96 compared with the serial implementation on Intel Xeon X5355 and Core Q6600.     

Solving the Shallow Water equations using 2D SPH particles for interactive applications

In this paper, we introduce a 2D particle-based approach to achieve realistic water surface behaviors for interactive applications. We formulate 2D particle-based Shallow Water equations using the Smoothed Particle Hydrodynamics. Particles defined with specific amount of water volume interplay with each other, which generates the horizon flow and the water surface motion. By the application of the particle-based Lagrangian framework to the 2D Shallow Water simulation, our method allows the water particles to move freely without being confined to a grid. The motion of the particles can represent global flow with dynamic waves covering a large area while avoiding extensive 3D fluid dynamics computation. The 2D particle-based Shallow Water equations are straightforward and computed fast with the GPU-based implementation. Experiments on a standard hardware demonstrate the performance of our approach which is running on the GPU, and the results show a realistic motion of the water surface at interactive rates.     

GPU-based high-performance computing for integrated surface–sub-surface flow modeling

The widespread availability of high-resolution lidar data provides an opportunity to capture micro-topographic control on the partitioning and transport of water for incorporation in coupled surface – sub-surface flow modeling. However, large-scale simulations of integrated flow at the lidar data resolution are computationally expensive due to the density of the computational grid and the iterative nature of the algorithms for solving nonlinearity. Here we present a distributed physically based integrated flow model that couples two-dimensional overland flow and three-dimensional variably saturated sub-surface flow on a GPU-based (Graphic Processing Unit) parallel computing architecture. Alternating Direction Implicit (ADI) scheme modified for GPU structure is used for numerical solutions in both models. Boundary condition switching approach is applied to partition potential water fluxes into actual fluxes for the coupling between surface and sub-surface models. The algorithms are verified using five benchmark problems that have been widely adopted in literature. This is followed by a large-scale simulation using lidar data. We demonstrate that the method is computationally efficient and produces physically consistent solutions. This computational efficiency suggests the feasibility of GPU computing for fully distributed, physics-based hydrologic models over large areas.     

Compression and rendering of iso-surfaces and point sampled geometry

In this paper we present a streaming compression scheme for gigantic point sets including per-point normals. This scheme extends on our previous Duodecim approach [21] in two different ways. First, we show how to use this approach for the compression and rendering of high-resolution iso-surfaces in volumetric data sets. Second, we use deferred shading of point primitives to considerably improve rendering quality. Iso-surface reconstruction is performed in a hexagonal close packing (HCP) grid, into which the initial data set is resampled. Normals are resampled from the initial domain using volumetric gradients. By incremental encoding, only slightly more than 3 bits per surface point and 5 bits per surface normal are required at high fidelity. The compressed data stream can be decoded in the graphics processing unit (GPU). Decoded point positions are saved in graphics memory, and they are then used on the GPU again to render point primitives. In this way high quality gigantic data sets can directly be rendered from their compressed representation in local GPU memory at interactive frame rates (see Fig. 1).     

Simulation and visualization of the Saint-Venant system using GPUs

We consider three high-resolution schemes for computing shallow-water waves as described by the Saint-Venant system and discuss how to develop highly efficient implementations using graphical processing units (GPUs). The schemes are well-balanced for lake-at-rest problems, handle dry states, and support linear friction models. The first two schemes handle dry states by switching variables in the reconstruction step, so that bilinear reconstructions are computed using physical variables for small water depths and conserved variables elsewhere. In the third scheme, reconstructed slopes are modified in cells containing dry zones to ensure non-negative values at integration points. We discuss how single and double-precision arithmetics affect accuracy and efficiency, scalability and resource utilization for our implementations, and demonstrate that all three schemes map very well to current GPU hardware. We have also implemented direct and close-to-photo-realistic visualization of simulation results on the GPU, giving visual simulations with interactive speeds for reasonably-sized grids.