期刊文献+

高性能GPU模拟器驱动设计研究 被引量:4

The driver design reserch for high-performance GPU simulater
下载PDF
导出
摘要 在研究GPU驱动设计时,考虑到驱动的开发量较大,提出了一种基于Mesa开源驱动框架的GPU模拟器驱动设计方法,在Mesa开源驱动的开发框架下,实现了一整套高性能GPU模拟器的驱动设计。本设计可适配OpenGL等多款应用程序接口(API),同时基于开源框架,大幅减小了开发难度。为了适配GPU模拟器的可编程性,在驱动设计中还集成了GPU的编译器设计,可将GLSL等编程语言转换为GPU中着色器的汇编代码,由着色器进行计算。根据模拟器驱动设计,提出了一套高性能GPU模拟器的接口设计,为模拟器各个模块提供了一套可读的驱动接口,指导了模拟器的结构设计。 The GPU driver design are studied.Considering the large amount of driver development,a design method of high-performance GPU simulator driver based on Mesa open source driver framework is proposed.Under the development framework of Mesa open source driver,a complete set of GPU simulator driver is realized.This design can be adapted to a variety of application programming interfaces(APIs),such as OpenGL.Based on the open source framework,the development difficulty is greatly reduced.In order to adapt to the programmability of the high-performance GPU simulator,the GPU compiler design is also integrated in the driver design,and the programming language,such as GLSL,can be changed into the assembly code of the shader in the GPU,which is calculated by the shader.According to the simulator driver design,a set of interface design of high-performance GPU simulator is proposed,which provides a set of readable drive interface for each module of the simulator and guides the structural design of the simulator.
作者 赵士彭 张立志 赵皓宇 苏孟豪 刘苏 Zhao Shipeng;Zhang Lizhi;Zhao Haoyu;Su Menghao;Liu Su(State Key Laboratory of Computer Architecture,Institute of Computer Technology,Chinese Academy of Sciences,Beijing 100190;Institute of Computing Technology,Chinese Academy of Sciences,Beijing 100190;University of Chinese Academy of Sciences,Beijing 100049;School of Computer Science and Technology,University of Science and Technology of China,Hefei 230026;Loongson Technology Corporation Limited,Beijing 100190)
出处 《高技术通讯》 EI CAS 北大核心 2020年第5期435-442,共8页 Chinese High Technology Letters
基金 国家自然科学基金(61521092,61432016) 中国科学院重点部署(ZDRW-XH-2017-1)资助项目。
关键词 开源驱动 应用程序接口(API) 驱动设计 GPU模拟器 模拟器接口 open source driver application programming interfaces(API) driver design GPU simulator simulator interface
  • 相关文献

参考文献2

二级参考文献28

  • 1http://ati. amd. com/technology/streamcomputing/product firestream_9270. html.
  • 2Kirk D. Nvidia Cuda Software and GPU Parallel Computing Architecture[C]//Proc of the 6th International Symposium on Memory Management, 2007 :103-104.
  • 3http://ati. amd. com/technology/streamcomputing/AMDBrook- plus. pelf.
  • 4Open Computing Language[EB/OL]. [2009-11-16]. http:// www. khronos. org/.
  • 5Luebke D, Harris M, Kruger J, et al. Gpgpu: General Pur- pose Computation on Graphics Hardware[C]//Proc of SIG- GRAPH' 04,2004 : 33.
  • 6Harris M. Mapping Computational Concepts to GPUs[C]// Proc of SIGGRAPH ' 05,2005 : 50.
  • 7The AMD Fusion Family of APUs[EB/OL]. [2009-11-15]. http://www, fusion, amd. com/.
  • 8Glaskowsky P N. NVIDIAs Fermi: The First Complete GPU Computing Architecture[R]. A White Paper Prepared Under Contract with NVIDIA Corporation, 2009: 1-26.
  • 9Seiler L, Carmean D, Sprangle E, et al. Larrabee: A Many- Core x86 Architecture for Visual Computing[J]. ACM Tran- seation on Graphics,2008,27(3) :11-15.
  • 10Pike A. DirectX 8 Tutorial[M]. 2006.

共引文献20

同被引文献12

引证文献4

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部