Yield enhancement of complex chips
View Patent ↗A graphics processing chip which includes parallel texturing pipelines, with task allocation units which can bypass inoperative ones of said pipelines. Chips which have some but not all pipelines operative can still have full functionality, although performance is reduced.
1. A graphics processor, comprising:
a plurality of parallellized graphics computational units; and
one or more task allocation units programmed to bypass defective ones of said units within said groups, and to distribute incoming tasks only among operative ones of said units.
2. The graphics processor of claim 1 , wherein each of said parallellized graphics computational units also includes respective multiple vertex processors.
3. The graphics processor of claim 1 , wherein each of said parallellized graphics computational units also includes respective texturing pipelines.
4. The graphics processor of claim 1 , wherein each of said parallellized graphics computational units also includes a respective memory controller.
5. A method of 3D graphics rendering which comprises: using a task allocation unit and parallellized graphics computational units with relations as recited in claim 1 .
6. The graphics processor of claim 1 , wherein one or more of said parallellized graphics computational units operate with no more than 4 operative vertex processors.
7. The graphics processor of claim 1 , wherein one or more of said parallellized graphics computational units operate with no more than 4 operative texturing pipelines.
8. The graphics processor of claim 3 , wherein said texturing pipelines also include a shading unit and a texture filter unit.
9. The graphics processor of claim 3 , wherein said texturing pipelines also include a shading unit and a primary texture cache.
10. A method of 3D graphics rendering, comprising the actions of:
providing a plurality of parallellized graphics computational units;
bypassing defective ones of said units, and
distributing incoming tasks only among operative ones of said units.
11. The method of claim 10 , wherein each of said parallellized graphics computational units also includes respective multiple vertex processors.
12. The method of claim 10 , wherein one or more of said parallellized graphics computational units operate with no more than 4 operative vertex processors.
13. The method of claim 10 , wherein each of said parallellized graphics computational units also includes respective multiple texturing pipelines.
14. The method of claim 10 , wherein one or more of said parallellized graphics computational units operate with no more than 4 operative texturing pipelines.
15. The method of claim 10 , wherein each of said parallellized graphics computational units also includes a respective memory controller.
16. The method of claim 13 , wherein said texturing pipelines also include a shading unit and a texture filter unit.
17. The method of claim 13 , wherein said texturing pipelines also include a shading unit and a primary texture cache.
18. A computer graphics system comprising:
means for providing a plurality of parallellized graphics computational units;
means for bypassing defective ones of said units, and
means for distributing incoming tasks only among operative ones of said units.
19. The system of claim 18 , wherein each of said parallellized graphics computational units also includes respective multiple vertex processors.
20. The system of claim 18 , wherein one or more of said parallellized graphics computational units operate with no more than 4 operative vertex processors.
21. The system of claim 18 , wherein each of said parallellized graphics computational units also includes respective multiple texturing pipelines.
22. The system of claim 18 , wherein one or more of said parallellized graphics computational units operate with no more than 4 operative texturing pipelines.
23. The system of claim 18 , wherein each of said parallellized graphics computational units also includes a respective memory controller.
24. The system of claim 21 , wherein said texturing pipelines also include a shading unit and a texture filter unit.
25. The system of claim 21 , wherein said texturing pipelines also include a shading unit and a primary texture cache.
26. A method for computer graphics system operation, comprising the actions of:
providing a plurality of parallellized rendering units;
bypassing defective ones of said units, and
distributing incoming tasks only among operative ones of said units.
27. The method of claim 26 , wherein each of said parallellized rendering units also includes respective multiple vertex processors.
28. The method of claim 26 , wherein one or more of said parallellized rendering units operate with no more than 4 operative vertex processors.
29. The method of claim 26 , wherein each of said parallellized rendering units also includes respective multiple texturing pipelines.
30. The method of claim 26 , wherein one or more of said parallellized rendering units operate with no more than 4 operative texturing pipelines.
31. The method of claim 26 , wherein each of said parallellized rendering units also includes a respective memory controller.
32. The method of claim 29 , wherein said texturing pipelines also include a shading unit and a texture filter unit.
33. The method of claim 29 , wherein said texturing pipelines also include a shading unit and a primary texture cache.