Graphics rendering systems
First Claim
1. A graphics rendering system, comprising:
- a framestore providing an array of pixel locations;
means for supplying a set of rendering instructions, each related to a graphical shape to be rendered in the framestore and each comprising an opcode indicative of a type of rendering operations, parameters to be used in that rendering operation, and an ordering code;
a plurality of processors operating in parallel to receive such rendering instructions and to write pixels in the framestore to render the related graphical shape; and
an ordering buffer having an array of memory cells each corresponding to a respective pixel location in the framestore;
wherein each processor, when determining whether to write a pixel at a pixel location in the framestore, is configured to write that respective pixel if and only if the ordering code of the graphical shape being rendered is more significant than the content of the memory cell in the buffer corresponding to that pixel, and when writing said pixel, to update that memory cell with the ordering code of the graphical shape being rendered.
6 Assignments
0 Petitions
Accused Products
Abstract
In a 2-D graphics rendering system having a plurality of processors (PROC 0 to PROC 3) which receive instructions from a common instruction register (IR) and render polygons in a framestore (FS), in order to permit asynchronous performance of the instructions and yet ensure that overlapping polygons are properly rendered, each instruction includes an ordering code, and before writing a pixel to the framestore (FS) each processor checks that the ordering code of the polygon it is rendering is more significant than an ordering code for that pixel stored in an ordering buffer in which case the pixel is written to the framestore (FS) and the order buffer (OB) is updated, but if not the pixel is not written.
-
Citations
5 Claims
-
1. A graphics rendering system, comprising:
-
a framestore providing an array of pixel locations; means for supplying a set of rendering instructions, each related to a graphical shape to be rendered in the framestore and each comprising an opcode indicative of a type of rendering operations, parameters to be used in that rendering operation, and an ordering code; a plurality of processors operating in parallel to receive such rendering instructions and to write pixels in the framestore to render the related graphical shape; and an ordering buffer having an array of memory cells each corresponding to a respective pixel location in the framestore; wherein each processor, when determining whether to write a pixel at a pixel location in the framestore, is configured to write that respective pixel if and only if the ordering code of the graphical shape being rendered is more significant than the content of the memory cell in the buffer corresponding to that pixel, and when writing said pixel, to update that memory cell with the ordering code of the graphical shape being rendered. - View Dependent Claims (2, 3, 4, 5)
-
Specification
-
- Resources
-
Current AssigneeZiiLabs Inc., Ltd. (Creative Technology Ltd.)
-
Original Assignee3D Labs
-
InventorsTrevett, Neil F.
-
Primary Examiner(s)Powell, Mark R.
-
Assistant Examiner(s)Tung, Kee M.
-
Application NumberUS07/996,911Time in Patent Office1,023 DaysField of Search395/157-158, 395/162-166, 395/425, 345/119-120, 345/185, 345/201, 345/203US Class Current345/505CPC Class CodesG06T 1/20 Processor architectures; Pr...
