A 18.4M Triangles/s 122.6 mW Tile Co-Processor for Embedded GPU Systems

Article Preview

Abstract:

This paper presents an efficient and accurate tile co-processor architecture which can be used in the tile based rendering systems. The design involves two key components, the vertex processing unit and the triangle tiling unit. The former part is used to get the vertices transformed, clipped and projected to generate the triangle list which located in the view frustum while the latter one reads in the triangle data and determines the tile list which indicates tiles that each triangle covers. A modified Bounding BOX (BBOX) test pipeline and a mask screening technology for different overlap types is proposed and employed in the design in order to get faster triangle binning with lower power consumption. The proposed architecture works at the frequency of 270 MHz, gains 18.4 M triangles tiling/sec with a power consumption less than 122.6 mW. The chip is implemented in 0.13 um CMOS technology and consumes 2.5 x 2.5 mm2 totally.

You might also be interested in these eBooks

Info:

Periodical:

Pages:

1050-1054

Citation:

Online since:

November 2013

Export:

Price:

Permissions CCC:

Permissions PLS:

Сopyright:

© 2014 Trans Tech Publications Ltd. All Rights Reserved

Share:

Citation:

* - Corresponding Author

[1] Imagination Technologies Ltd, POWERVR Series5 Graphics SGX Architecture Guide for Developers, 1. 0. 8 Ed., Imagination Technologies Ltd, (2011).

Google Scholar

[2] T. Akenine-Moller, E. Haines, 3rd Ed., Realtime Rendering, A K Peters/CRC Press, Florida, (2008).

Google Scholar

[3] S. O. Woo et al., Fast Tile-binning Method by detecting 1D-Overlapped Primitives, Consumer Electronics (ICCE), 2013 IEEE International Conference on, Las Vegas, NV, pp.147-148, Jan. (2013).

DOI: 10.1109/icce.2013.6486834

Google Scholar

[4] H. Hsieh, C. Hsiao, Methods for precise false-overlap detection in tile-based rendering", In proceeding of: Computational Science and Engineering, 2009. CSE , 09. International Conference on, Volume: 2, 2009, Vancouver, BC, pp.414-419, Aug. (2009).

DOI: 10.1109/cse.2009.466

Google Scholar

[5] I. Antochi et al., Efficient Tile-Aware Bounding-Box Overlap Test for Tile-based Rendering, System-on-Chip, 2004. Proceedings. 2004 International Symposium on, pp.165-168, Nov. (2004).

DOI: 10.1109/issoc.2004.1411177

Google Scholar

[6] Imagination Tech. Corp, PowerVR. 3D Graphical Processing Tile Based Rendering-The Future of 3D , white paper, Imagination Tech. Corp, (2000).

Google Scholar

[7] I. Antochi et al., Scene Management Models and Overlap Tests for Tile-Based Rendering, In Proc. EUROMICRO Symp. on Digital System Design (DSD 2004) , pp.424-431, 31 Aug. -3 Sept. (2004).

DOI: 10.1109/dsd.2004.1333306

Google Scholar

[8] K. Karagianni et al. : IEEE Transactions on Computers, vol. 50, no. 6, p.609–622, June (2001).

Google Scholar

[9] L. Gao et al. A low area clipping engine in 3d graphics system, CCWMC, p.491–495, (2011).

Google Scholar

[10] Imagination Technologies, MIPS-3D ASE, http: /www. imgtec. com/mips/mips-3d-ase. asp.

Google Scholar