Fast constructive-solid geometry display in the pixel-powers graphics system
SIGGRAPH '86 Proceedings of the 13th annual conference on Computer graphics and interactive techniques
Advances in computer graphics hardware II
A characterization of ten rasterization techniques
SIGGRAPH '89 Proceedings of the 16th annual conference on Computer graphics and interactive techniques
PROOF: an architecture for rendering in object space
Advances in computer graphics hardware III
The triangle processor and normal vector shader: a VLSI system for high performance graphics
SIGGRAPH '88 Proceedings of the 15th annual conference on Computer graphics and interactive techniques
Subanosecond pixel rendering with million transistor chips
SIGGRAPH '88 Proceedings of the 15th annual conference on Computer graphics and interactive techniques
A display system for the Stellar graphics supercomputer model GS1000
SIGGRAPH '88 Proceedings of the 15th annual conference on Computer graphics and interactive techniques
The Silicon Graphics 4D/240GTX Superworkstation
IEEE Computer Graphics and Applications
Parallel processing image synthesis and anti-aliasing
SIGGRAPH '81 Proceedings of the 8th annual conference on Computer graphics and interactive techniques
Hi-index | 0.00 |
Current graphics processors are very slow for displaying shaded 3D objects. A lot of work is being done in order to define faster display processors by using massive parallelism and VLSI components. Our proposal goes along this line with the supplementary aim of displaying images in real time, i.e., 25 or 30 times per second. We choose to design a graphics module without any working memory and thus without frame buffer. A massive parallelism over objects, and thus a pixel pipe-line, are used. Each Object Processor handles one 3D object; all the processors work in a synchronous way, processing the same pixel simultaneously at pixel rate. These processors are built from very simple Elementary Processors (2 adders, 2 registers and 6 memory words) computing linear or quadratic expressions V(x,y), where (x,y) are the coordinates of a pixel. A pipelined tree made of basic operators (min, max, or, and, ... ) gathers the results given by the Object Processors and makes inter-objects operations, at least hidden part elimination. Such a choice of course involves a high hardware complexity when displaying rather simple scenes. However, we feel that it is the price to pay for building graphics processors allowing real-time interactive animation (e.g., the graphics unit of a driving simulator).