Hardware implementation of micropolygon rasterization with motion and defocus blur

John Brunhaver, K. Fatahalian, P. Hanrahan

Research output: Chapter in Book/Report/Conference proceedingConference contribution

13 Citations (Scopus)

Abstract

Current GPUs rasterize micropolygons (polygons approximately one pixel in size) inefficiently. Additionally, they do not natively support triangle rasterization with jittered sampling, defocus, or motion blur. We perform a microarchitectural study of fixed-function micropolygon rasterization using custom circuits. We present three rasterization designs: the first optimized for triangle micropolygons that are not blurred, a second for stochastic rasterization of micropolygons with motion and defocus blur, and third that is a hybrid combination of the two. Our designs achieve high area and power efficiency by using low-precision operations and rasterizing pairs of adjacent triangles in parallel. We demonstrate optimized designs synthesized in a 45 nm process showing that a micropolygon rasterization unit with a throughput of 3 billion micropolygons per second would consume 2.9 W and occupy 4.1 mm2 which is 0.77% of the die area of a GeForce GTX 480 GPU. Categories and Subject Descriptors (according to ACM CCS): I.3.1 [Computer Graphics]: Graphics processors-I.3.7 Three-Dimensional Graphics and Realism.

Original languageEnglish (US)
Title of host publicationHigh-Performance Graphics 2010 - ACM SIGGRAPH / Eurographics Symposium Proceedings, HPG 2010
EditorsSamuli Laine, Warren Hunt, Michael Doggett
PublisherAssociation for Computing Machinery
Pages1-9
Number of pages9
ISBN (Print)9783905674262
StatePublished - Jun 25 2010
Externally publishedYes
Event2nd ACM SIGGRAPH / Eurographics on High-Performance Graphics, HPG 2010 - Saarbrucken, Germany
Duration: Jun 25 2010Jun 27 2010

Other

Other2nd ACM SIGGRAPH / Eurographics on High-Performance Graphics, HPG 2010
CountryGermany
CitySaarbrucken
Period6/25/106/27/10

Fingerprint

Hardware
Computer graphics
Pixels
Throughput
Rasterization
Sampling
Networks (circuits)
Graphics processing unit

ASJC Scopus subject areas

  • Hardware and Architecture
  • Software
  • Computational Theory and Mathematics
  • Computer Graphics and Computer-Aided Design

Cite this

Brunhaver, J., Fatahalian, K., & Hanrahan, P. (2010). Hardware implementation of micropolygon rasterization with motion and defocus blur. In S. Laine, W. Hunt, & M. Doggett (Eds.), High-Performance Graphics 2010 - ACM SIGGRAPH / Eurographics Symposium Proceedings, HPG 2010 (pp. 1-9). Association for Computing Machinery.

Hardware implementation of micropolygon rasterization with motion and defocus blur. / Brunhaver, John; Fatahalian, K.; Hanrahan, P.

High-Performance Graphics 2010 - ACM SIGGRAPH / Eurographics Symposium Proceedings, HPG 2010. ed. / Samuli Laine; Warren Hunt; Michael Doggett. Association for Computing Machinery, 2010. p. 1-9.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Brunhaver, J, Fatahalian, K & Hanrahan, P 2010, Hardware implementation of micropolygon rasterization with motion and defocus blur. in S Laine, W Hunt & M Doggett (eds), High-Performance Graphics 2010 - ACM SIGGRAPH / Eurographics Symposium Proceedings, HPG 2010. Association for Computing Machinery, pp. 1-9, 2nd ACM SIGGRAPH / Eurographics on High-Performance Graphics, HPG 2010, Saarbrucken, Germany, 6/25/10.
Brunhaver J, Fatahalian K, Hanrahan P. Hardware implementation of micropolygon rasterization with motion and defocus blur. In Laine S, Hunt W, Doggett M, editors, High-Performance Graphics 2010 - ACM SIGGRAPH / Eurographics Symposium Proceedings, HPG 2010. Association for Computing Machinery. 2010. p. 1-9
Brunhaver, John ; Fatahalian, K. ; Hanrahan, P. / Hardware implementation of micropolygon rasterization with motion and defocus blur. High-Performance Graphics 2010 - ACM SIGGRAPH / Eurographics Symposium Proceedings, HPG 2010. editor / Samuli Laine ; Warren Hunt ; Michael Doggett. Association for Computing Machinery, 2010. pp. 1-9
@inproceedings{ffdfa268cc88436e8b13e47387479514,
title = "Hardware implementation of micropolygon rasterization with motion and defocus blur",
abstract = "Current GPUs rasterize micropolygons (polygons approximately one pixel in size) inefficiently. Additionally, they do not natively support triangle rasterization with jittered sampling, defocus, or motion blur. We perform a microarchitectural study of fixed-function micropolygon rasterization using custom circuits. We present three rasterization designs: the first optimized for triangle micropolygons that are not blurred, a second for stochastic rasterization of micropolygons with motion and defocus blur, and third that is a hybrid combination of the two. Our designs achieve high area and power efficiency by using low-precision operations and rasterizing pairs of adjacent triangles in parallel. We demonstrate optimized designs synthesized in a 45 nm process showing that a micropolygon rasterization unit with a throughput of 3 billion micropolygons per second would consume 2.9 W and occupy 4.1 mm2 which is 0.77{\%} of the die area of a GeForce GTX 480 GPU. Categories and Subject Descriptors (according to ACM CCS): I.3.1 [Computer Graphics]: Graphics processors-I.3.7 Three-Dimensional Graphics and Realism.",
author = "John Brunhaver and K. Fatahalian and P. Hanrahan",
year = "2010",
month = "6",
day = "25",
language = "English (US)",
isbn = "9783905674262",
pages = "1--9",
editor = "Samuli Laine and Warren Hunt and Michael Doggett",
booktitle = "High-Performance Graphics 2010 - ACM SIGGRAPH / Eurographics Symposium Proceedings, HPG 2010",
publisher = "Association for Computing Machinery",

}

TY - GEN

T1 - Hardware implementation of micropolygon rasterization with motion and defocus blur

AU - Brunhaver, John

AU - Fatahalian, K.

AU - Hanrahan, P.

PY - 2010/6/25

Y1 - 2010/6/25

N2 - Current GPUs rasterize micropolygons (polygons approximately one pixel in size) inefficiently. Additionally, they do not natively support triangle rasterization with jittered sampling, defocus, or motion blur. We perform a microarchitectural study of fixed-function micropolygon rasterization using custom circuits. We present three rasterization designs: the first optimized for triangle micropolygons that are not blurred, a second for stochastic rasterization of micropolygons with motion and defocus blur, and third that is a hybrid combination of the two. Our designs achieve high area and power efficiency by using low-precision operations and rasterizing pairs of adjacent triangles in parallel. We demonstrate optimized designs synthesized in a 45 nm process showing that a micropolygon rasterization unit with a throughput of 3 billion micropolygons per second would consume 2.9 W and occupy 4.1 mm2 which is 0.77% of the die area of a GeForce GTX 480 GPU. Categories and Subject Descriptors (according to ACM CCS): I.3.1 [Computer Graphics]: Graphics processors-I.3.7 Three-Dimensional Graphics and Realism.

AB - Current GPUs rasterize micropolygons (polygons approximately one pixel in size) inefficiently. Additionally, they do not natively support triangle rasterization with jittered sampling, defocus, or motion blur. We perform a microarchitectural study of fixed-function micropolygon rasterization using custom circuits. We present three rasterization designs: the first optimized for triangle micropolygons that are not blurred, a second for stochastic rasterization of micropolygons with motion and defocus blur, and third that is a hybrid combination of the two. Our designs achieve high area and power efficiency by using low-precision operations and rasterizing pairs of adjacent triangles in parallel. We demonstrate optimized designs synthesized in a 45 nm process showing that a micropolygon rasterization unit with a throughput of 3 billion micropolygons per second would consume 2.9 W and occupy 4.1 mm2 which is 0.77% of the die area of a GeForce GTX 480 GPU. Categories and Subject Descriptors (according to ACM CCS): I.3.1 [Computer Graphics]: Graphics processors-I.3.7 Three-Dimensional Graphics and Realism.

UR - http://www.scopus.com/inward/record.url?scp=84874186913&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84874186913&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84874186913

SN - 9783905674262

SP - 1

EP - 9

BT - High-Performance Graphics 2010 - ACM SIGGRAPH / Eurographics Symposium Proceedings, HPG 2010

A2 - Laine, Samuli

A2 - Hunt, Warren

A2 - Doggett, Michael

PB - Association for Computing Machinery

ER -