Computer graphics system utilizing parallel processing for enhanced performance

• US 5,821,950 A
• Filed: 04/18/1996
• Issued: 10/13/1998
• Est. Priority Date: 04/18/1996
• Status: Expired due to Term

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Company (HP Inc.)
• Inventors: Handgen, Erin A., Matthews, Mary A., Rentschler, Eric M., Krech, Alan S. Jr., Shah, Monish S.
• Primary Examiner(s): Bayerl, Raymond J.
• Assistant Examiner(s): CHAUHAN, ULKA J
• Application Number:
  US08/634,458
• Time in Patent Office: 908 Days
• Field of Search: 395/118-139, 395/501, 395/502, 395/503, 395/505, 395/511, 395/522
• US Class Current: 345/505
• CPC Class Codes: G06T 1/20 Processor architectures; Processor configuration, e.g. pipelining
• CPC Class Codes: G06T 11/40 Filling a planar surface by adding surface attributes, e.g. colour or texture
• CPC Class Codes: G06T 15/005 General purpose rendering architectures

Abstract

A computer graphics system includes a plurality of geometry accelerators for processing vertex data representative of graphics primitives and providing rendering data. The system includes a distributor responsive to a stream of vertex data for distributing to the geometry accelerators chunks of the vertex data for processing by the geometry accelerators to provide chunks of rendering data. The distributor generates an end of chunk bit indicative of the end of each of the chunks of vertex data. The system further includes a concentrator for receiving the chunks of rendering data from each of the geometry accelerators and for combining the chunks of rendering data into a stream of rendering data in response to end of chunk bits. The stream of rendering data and the stream of vertex data represent sequences of graphics primitives having the same order. A rasterizer generates pixel data representative of a graphics display in response to the stream of rendering data.

Claims

1. A computer graphics system comprising:

a plurality of geometry accelerators for processing vertex data representative of graphics primitives and providing rendering data, each of said plurality of geometry accelerators having an input and an output;

a distributor responsive to a stream of vertex data for distributing to said plurality of geometry accelerators chunks of said vertex data for processing by said plurality of geometry accelerators to provide chunks of rendering data, said distributor including means for generating, for each of said chunks of vertex data, an end of chunk bit indicative of the end of a corresponding chunk of vertex data and means for distributing each of said end of chunk bits to one of said geometry accelerators with said corresponding chunk of vertex data, each of said geometry accelerators including means for transmitting each of said end of chunk bits from said input to said output;

a concentrator for receiving said chunks of rendering data and said end of chunk bits from each of said geometry accelerators and responsive to said end of chunk bits for combining said chunks of rendering data from each of said geometry accelerators into a stream of rendering data, wherein said stream of rendering data and said stream of vertex data represent sequences of graphics primitives having the same order; and

a rasterizer responsive to said stream of rendering data for generating pixel data representative of a graphics display.

2. A computer graphics system as defined in claim 1 wherein said concentrator comprises a plurality of first-in first-out (FIFO) buffers respectively connected to outputs of said geometry accelerators, a multiplexer having inputs respectively coupled to the outputs of said FIFO buffers and an output for providing said stream of rendering data, and a multiplexer controller responsive to said end of chunk bits for sequentially selecting outputs of said FIFO buffers in accordance with a defined sequence.

3. A computer graphics system as defined in claim 1 wherein said distributor includes a configuration register for defining a number of geometry accelerators in the system and for defining a sequence for distributing said chunks of vertex data to said plurality of geometry accelerators, said sequence depending on the number of geometry accelerators in the system, and means for distributing said chunks of vertex data to said plurality of geometry accelerators in said defined sequence.

4. A computer graphics system as defined in claim 3 wherein said concentrator includes means combining said chunks of rendering data from said plurality of geometry accelerators in said defined sequence.

5. A computer graphics system as defined in claim 1 wherein each of said geometry accelerators includes means for maintaining said end of chunk bit in said vertex data even if a portion of said vertex data is clipped during processing.

6. A computer graphics system as defined in claim 1 including at least three geometry accelerators.

7. A computer graphics system as defined in claim 1 including first, second and third geometry accelerators, wherein said chunks of vertex data are coupled from said distributor to said first and second geometry accelerators on a first distributor bus, said chunks of vertex data are coupled from said distributor to said third geometry accelerator on a second distributor bus, said chunks of rendering data are coupled from said first and second geometry accelerators to said concentrator on a first concentrator bus, and said chunks of rendering data are coupled from said third geometry accelerator to said concentrator on a second concentrator bus.

8. A computer graphics system as defined in claim 1 including first, second, third and fourth geometry accelerators, wherein said chunks of vertex data are coupled from said distributor to said first and second geometry accelerators on a first distributor bus, said chunks of vertex data are coupled from said distributor to said third and fourth geometry accelerators on a second distributor bus, said chunks of rendering data are coupled from said first and second geometry accelerators to said concentrator on a first concentrator bus and said chunks of rendering data are coupled from said third and fourth geometry accelerators to said concentrator on a second concentrator bus.

9. A computer graphics system as defined in claim 1 wherein said vertex data is coupled to said geometry accelerators on first and second distributor buses and wherein said distributor includes a first FIFO buffer for supplying said chunks of vertex data on said first distributor bus, a second FIFO buffer for supplying said chunks of vertex data on said second distributor bus and a distributor for distributing said chunks of vertex data between said first and second FIFO buffers.

10. A computer graphics system as defined in claim 1 wherein each of said end of chunk bits is associated with a last word of each of said chunks of vertex data.

11. A method for processing vertex data in a computer graphics system, comprising the steps of:

distributing chunks of vertex data representative of graphics primitives to a plurality of geometry accelerators, each having an input and an output;

generating, for each of said chunks of vertex data, an end of chunk bit indicative of the end of a corresponding chunk of vertex data;

distributing each of said end of chunk bits to one of said geometry accelerators with said corresponding chunk of vertex data;

said plurality of geometry accelerators processing said chunks of vertex data to provide chunks of rendering data;

said plurality of geometry accelerators transmitting each of said end of chunk bits from said input to said output;

combining said chunks of rendering data in response to said end of chunk bits to provide a stream of rendering data, wherein said stream of rendering data and said stream of vertex data represent sequences of graphics primitives having the same order; and

rasterizing said stream of rendering data to provide pixel data representative of a graphics display.

12. A method for processing vertex data as defined in claim 11 further including the step of defining a sequence for distributing said chunks of vertex data to said plurality of geometry accelerators and wherein said chunks of vertex data are distributed to said plurality of geometry accelerators in said defined sequence.

13. A method for processing vertex data as defined in claim 12 wherein said chunks of rendering data are combined in said defined sequence.

14. A method for processing vertex data as defined in claim 11 further including the step of maintaining said end of chunk bits in said vertex data even if a portion of said vertex data is clipped during processing.