Browse Prior Art Database

Precisely-Ordered Parallel 3D Polygon Rendering Algorithm

IP.com Disclosure Number: IPCOM000123573D
Original Publication Date: 1999-Jan-01
Included in the Prior Art Database: 2005-Apr-05
Document File: 3 page(s) / 65K

Publishing Venue

IBM

Related People

Oie, M: AUTHOR [+2]

Abstract

Disclosed is a method for rendering 3D polygons on the 3D graphics controller, which enables multiple 3D polygons to be rendered in parallel while preserving precise rendering order, by interlocking multiple rendering pipelines according to the result of overlap analysis for each polygon.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 69% of the total text.

Precisely-Ordered Parallel 3D Polygon Rendering Algorithm

   Disclosed is a method for rendering 3D polygons on the 3D
graphics controller, which enables multiple 3D polygons to be
rendered in parallel while preserving precise rendering order, by
interlocking multiple rendering pipelines according to the result of
overlap analysis for each polygon.

   This method first analyses the overlap of each polygon,
before dispatching the given vertex data stream into each rendering
pipeline (FIG1).  In general, the vertex data stream for the polygon
is given in the form of the Triangle Strip or the Triangle Fan.  In
such cases, the overlap of each polygon may be determined according
to the outer product of the vectors that compose the polygon.

  For example, a Triangle Strip which generates two polygons is shown
below (FIG2).  In this example, it is determined that the triangle
(V1,V2,V3) does not have any overlap with the triangle (V2,V3,V4),
because the outer product of the vector V2V3 and the vector V2V1 is
positive, but the outer product of the vector V2V3 and the vector
V2V4 is negative.

   This method next dispatches the polygons into each pipeline
(FIG1).  According to the result of the overlap analysis described
above, it is determined that the triangles (V1,V2,V3) and (V2,V3,V4)
and the triangles (V2,V3,V4) and (V3,V4,V5) may be processed in
parallel by the two rendering pipelines, because they do not overlap
with each other.  However, the execution of the triang...