摘要
This paper studies some programming techniques for low power rendering for 3 D graphics. These techniques are derived from analysis and simulation results of hardware circuits of GPU. Although low power3 D graphics hardware design has been studied by other researchers,low power programming techniques from hardware perspective have not been investigated in depth. There are many factors that affect 3 D graphics rendering performance,such as the number of vertices,vertex sharing,level of details,texture mapping,and rendering algorithms. An analytical study of graphics rendering workload is performed and the effect of a number of programming tips such as vertex sharing,clock gating and buffering of unmoving or translational objects is deeply studied. The results presented in this paper can be used to guide 3 D graphics programming for optimizing both power consumption and performance.
This paper studies some programming techniques for low power rendering for 3 D graphics. These techniques are derived from analysis and simulation results of hardware circuits of GPU. Although low power3 D graphics hardware design has been studied by other researchers,low power programming techniques from hardware perspective have not been investigated in depth. There are many factors that affect 3 D graphics rendering performance,such as the number of vertices,vertex sharing,level of details,texture mapping,and rendering algorithms. An analytical study of graphics rendering workload is performed and the effect of a number of programming tips such as vertex sharing,clock gating and buffering of unmoving or translational objects is deeply studied. The results presented in this paper can be used to guide 3 D graphics programming for optimizing both power consumption and performance.
引文
[1] Chiueh Tzi-cker,Lin Wei-jen. Characterization of static3D graphics workloads. HWWS'97 Proceedings of the ACM SIGGRAPH/EUROGRAPHICS workshop on Graphics hardware. New York:ACM,1997. 17-24.DOI:10.1145/258694.258703.
[2] Mitra T,Chiueh T C. Dynamic 3D graphics workload characterization and the architectural implications.Proceedings of the 32nd ACM/IEEE Int Symp. On Microarchitecture(MICRO). Piscataway:IEEE,1999.62-71. DOI:10.1109/MICRO.1999.809444.
[3] Roca J, Moya V, Gonzalez C, et al. Workload characterization of 3D games, IEEE International Symposium on Workload Characterization. Piscataway:IEEE,2006.17-26. DOI:10.1109/IISWC.2006.302726.
[4]Mochocki B C,Lahiri K,Cadambi S,et al. Signaturebased workload estimation for mobile 3D graphics.Proceedings of the 43rd ACM/IEEE,Design Automation Conference. Piscataway:IEEE,2006.592-597. DOI:10.1145/1146909.1147062.
[5] Ma X,Deng Z,Dong M,et al. Characterizing the performance and power consumption of 3D mobile games.Computer,2012, 46(4):76-82. DOI:10.1109/MC.2012.190.
[6] Mochicki B,Lahiri K,Cadambi S. Power Analysis of mobile 3D graphics. In Proceedings of the Conference on Design,Automation and Test in Europe. Piscataway:IEEE, 2006. 502-507. DOI:10. 1109/DATE. 2006.243859.
[7] Nagasaka H,Maruyama N,Nukada A,et al. Statistical power modeling of GPU kernels using performance counters. Proceedings of the 2010 International Green Computing Conference(IGCC). Piscataway:IEEE,2010. 115-122. DOI:10. 1109/GREENCOMP. 2010.5598315.
[8] Ma X H,Dong M,Zhong L,et al. Statistical power consumption analysis and modeling for GPU-based computing. Proceedings of the Workshop on Power Aware Computing and Systems(HotPower). New York:ACM,2009.
[9] Adhinarayanan V,Subramaniam B,Feng W C,et al.Online Power estimation of graphics processing units.Proceedings of the 16th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing.Piscataway:IEEE, 2016. 245-254. DOI:10. 1109/CCGrid.2016.93.
[10]Kasichayanula K,Terpstra D,Luszczek P,et al. Power aware computing on GPUs. Proceedings of the 2012Symposium on Application Accelerators in High Performance Computing(SAAHPC). Piscataway:IEEE,2012. 64-73. DOI:10.1109/SAAHPC.2012.26.
[11]Wu G,Greathouse J L,Lyashevsky A,et al. GPGPU performance and power estimation using machine learning.2015 IEEE 21st International Symposium on High Performance Computer Architecture(HPCA).Piscataway:IEEE,2015. 564-576. DOI:10.1109/HPCA.2015.7056063.
[12]Collange S,Defour D,Tisserand A. Power consumption of GPUs from a software perspective. International Conference on Computational Science. Berlin:Springer,2009,5544:914-923. DOI:10.1007/978-3-642-01970-8_92.
[13]Rakvic R,Broussard R,Ngo H. Energy efficient iris recognition with graphics processing units. IEEE Biometrics Compendium. Pisacataway:IEEE Access,2016,4:2831-2839. DOI:10. 1109/ACCESS. 2016.2571747.
[14] NVIDIA Corp. NVIDIA Tegra 4 Family GPU Architecture(Whitepaper). http://www. nvidia. cn/object/white-papers-cn.html,February 2013.
[15]NVIDIA Corp. NVIDIATegraX1 NVIDIA’S New Mobile Superchip(Whitepaper). http://www.nvidia.cn/object/white-papers-cn.html,January 2015.
[16]Xing L D,Li T,Huang H,et al. Efficient modeling and analysis of energy consumption for 3D graphics rendering.Integration,the VLSI Journal,2016,55:455-464. DOI:10.1016/j.vlsi.2016.02.009.