Occlusion prediction graphics processing system and method
An occlusion prediction graphics processing system and method are presented in accordance with embodiments of the present invention. An occlusion prediction graphics processing method is utilized to predict which pixel values are eventually occluded before intermediate processing stages are performed on the pixel values. For example, occlusion results are predicted before the occlusion stage of a graphics pipeline. The occlusion prediction results are based upon an occlusion value received from later in a graphics processing pipeline (e.g., a raster operation stage). A convex polygonal prediction area can be established and a nearest vertex of the convex polygonal prediction area is selected for prediction analysis. Pixel values are removed or discarded from the pipeline based upon the occlusion prediction results and do not unnecessarily occupy processing resources. Removal of the pixel values from the pipeline includes pixels values associated with pixels in the convex polygonal prediction area. Pixel shading is performed on the remaining pixels.
Latest NVIDIA Corporation Patents:
- Extended through wafer vias for power delivery in face-to-face dies
- Just in time compilation using link time optimization
- Online fault detection in ReRAM-based AI/ML
- Intelligent thermosyphon system for datacenter cooling systems
- Rail power density aware standard cell placement for integrated circuits
The present invention relates to the field of graphics processing.
BACKGROUND OF THE INVENTIONElectronic systems and circuits have made a significant contribution towards the advancement of modern society and are utilized in a number of applications to achieve advantageous results. Numerous electronic technologies such as digital computers, calculators, audio devices, video equipment, and telephone systems facilitate increased productivity and cost reduction in analyzing and communicating data, ideas and trends in most areas of business, science, education and entertainment. Electronic systems designed to produce these results usually involve interfacing with a user and the interfacing often involves presenting graphical representation of images. Displaying graphics images usually requires intensive data processing which traditionally takes considerable time to process and often consumes significant power.
Computer graphics systems typically utilize a sequential stage or “pipeline” type process to map a three dimensional scene in the world coordinate system to a two dimensional projection (e.g., on a display screen). In most computer graphic systems an image is represented as a raster (an array) of logical picture elements (pixels). Pipelines typically assign parameter values to each pixel and the parameter values determine the nature of the projection on the display screen. The parameter values are digital values corresponding to certain attributes of the image (e.g. color, depth, etc.) measured over a small area of the image represented by a pixel. Typically each graphical image is represented by thousands of combined pixels. Providing information for each pixel is very data intensive and consumes a significant amount of processing resources.
There are a number of stages or processes included in a typical graphics pipeline. Various manipulations of pixel data are implemented at each stage in the pipeline. These manipulations often involve numerous computational processes that take a relatively long time to complete. In addition, the processing consumes significant power and can be a significant drain on limited power supplies, such as a battery. One process performed in a typical graphics pipeline is to eliminate pixel values that are occluded, such as values associated with “hidden” surfaces. The occlusion determining process typically occurs near or at the end of a graphics pipeline after a number of processing operations have been performed to establish a variety of pixel values. These processes are often performed even on pixel values that are eventually discarded at the end of the graphics pipeline.
The rate at which images are rendered in typical graphics systems is often critical to proper presentation of the information. Slow rendering rates often result in undesirable choppy or “jerky” presentations which usually results in a user experience that is non-immersive and unpleasant. The rate at which graphics systems can render images is often limited by the rate at which the processing devices can process the graphics information. However, user's tend to have ever increasing demands for ever more spectacular and clearer images with better resolutions. Achieving better resolution often involves more graphic information processing and advanced applications. As more information associated with sophisticated applications and complex image rending is fed into traditional graphics pipelines the time required to process all the information increases since the graphics processing capabilities of the graphics systems typically have an upper limit. In addition, accessing the increased amounts of information also increases the time involved in retrieving the information from various memories. The increases in time to perform processing and information accesses typically slows the rendering rate and adversely impacts the graphics presentation.
SUMMARYAn occlusion prediction graphics processing system and method are presented in accordance with embodiments of the present invention. In one embodiment, an occlusion prediction graphics processing method is utilized to predict which pixel values are eventually occluded and culling is executed before intermediate processing stages are performed on the pixel values. For example, occlusion results are predicted before the occlusion stage of a graphics pipeline. The occlusion prediction results are based upon an occlusion value received from later in a graphics processing pipeline (e.g., a raster operation stage). The occlusion value is received from a stage later in the graphics processing pipeline that already performs frame buffer accesses for its own purposes and additional memory access bandwidth is not consumed reading an occlusion value (e.g., from a frame buffer). In one embodiment a square or rectangle prediction area is established and a nearest vertex of the prediction area is selected for prediction analysis. Pixel values are removed or discarded from the pipeline based upon the occlusion prediction results and do not unnecessarily occupy processing resources. In one exemplary implementation, removal of the pixel values from the pipeline includes pixels values associated with pixels in the prediction area. Intermediate graphics processing pipeline stages (e.g., pixel shading) is performed on the remaining pixels.
The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments of the invention by way of example and not by way of limitation. The drawings referred to in this specification should be understood as not being drawn to scale except if specifically noted.
Reference will now be made in detail to the preferred embodiments of the invention, examples of which are illustrated in the accompanying drawings. While the invention will be described in conjunction with the preferred embodiments, it will be understood that they are not intended to limit the invention to these embodiments. On the contrary, the invention is intended to cover alternatives, modifications and equivalents, which may be included within the spirit and scope of the invention as defined by the appended claims. Furthermore, in the following detailed description of the present invention, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be obvious to one of ordinary skill in the art that the present invention may be practiced without these specific details. In other instances, well known methods, procedures, components, and circuits have not been described in detail as not to unnecessarily obscure aspects of the present invention.
Some portions of the detailed descriptions which follow are presented in terms of procedures, logic blocks, processing, and other symbolic representations of operations on data bits within a computer memory. These descriptions and representations are the means generally used by those skilled in data processing arts to effectively convey the substance of their work to others skilled in the art. A procedure, logic block, process, etc., is here, and generally, conceived to be a self-consistent sequence of steps or instructions leading to a desired result. The steps include physical manipulations of physical quantities. Usually, though not necessarily, these quantities take the form of electrical, magnetic, optical, or quantum signals capable of being stored, transferred, combined, compared, and otherwise manipulated in a computer system. It has proven convenient at times, principally for reasons of common usage, to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, or the like.
It should be borne in mind, however, that all of these and similar terms are associated with the appropriate physical quantities and are merely convenient labels applied to these quantities. Unless specifically stated otherwise as apparent from the following discussions, it is appreciated that throughout the present application, discussions utilizing terms such as “processing”, “computing”, “calculating”, “determining”, “displaying” or the like, refer to the action and processes of a computer system, or similar processing device (e.g., an electrical, optical, or quantum, computing device), that manipulates and transforms data represented as physical (e.g., electronic) quantities. The terms refer to actions and processes of the processing devices that manipulate or transform physical quantities within a computer system's component (e.g., registers, memories, other such information storage, transmission or display devices, etc.) into other data similarly represented as physical quantities within other components.
In one embodiment of the present invention a culling prediction is performed based upon a culling value received from a stage later in a graphics processing pipeline. The culling value is associated with a position in a viewing volume. The viewing volume is formed by a near plane closest to a user's perspective or “eye” (e.g., a closer Z value) and a farthest plane from a user's perspective or “eye” (e.g., a deeper Z value). References to near or nearer indicate a position within the view volume closer the user's perspective and far or farther indicate a position within the view volume a greater distance away from the user's perspective or “eye”.
In step 110, an occlusion prediction is performed based upon an occlusion value received from a stage later in a graphics processing pipeline. By receiving the occlusion value from a stage later in the graphics processing pipeline that already performs frame buffer accesses for its own purposes, additional memory access bandwidth is not consumed reading an occlusion value (e.g., from a frame buffer). In one embodiment of the present invention, the occlusion information is received from a stage (e.g., a raster operation stage) in the graphics processing pipeline that initiates final reads and writes of occlusion information (e.g., to a Z buffer, frame buffer, etc.). For example, an occlusion value is received from a graphics pipeline processing stage in which final depth (e.g., Z value) and stencil comparisons are performed. In one embodiment, the predicting is performed within a rasterizing stage based upon a culling value received from a raster operation (ROP) stage of a graphics processing pipeline.
In one embodiment of the present invention, the occlusion prediction is based upon an occlusion value associated with pixels within a prediction area. A quantity and configuration of pixels within bounds of a prediction area are established (e.g., 2 by 4 pixel prediction area, 4 by 4 pixel prediction area, etc.). If the occlusion value received from the end of the graphics processing pipeline indicates that the pixels in the prediction area currently being rasterized are behind the occlusion value, the prediction is the pixels currently being rasterized should be culled.
In one exemplary implementation, the prediction area is a square or rectangle. For occlusion testing, the portion of the prediction area that overlaps the primitive (e.g., a triangle) being drawn is considered. The overlap is convex and permits the nearest depth value within the overlap region to be ascertained. Since depth varies linearly across the primitive the nearest depth value will be found at a vertex (e.g., corner) of the overlap region. The vertex can be a corner of the square or rectangular prediction area, or at a vertex of the original primitive, or at an intersection of a prediction area side with a primitive side. The nearest vertex can be also determined using the primitive's depth plane equation. If the nearest or closest vertex has a depth value that is farther back or away from a user's eye, (e.g., has a deeper Z value) than an occlusion value (e.g., corresponding to the depth value bound on the already drawn pixels within the prediction area), a prediction is made that the primitive's depth values within the prediction area presently in raster stage are occluded by the corresponding pixel value in the depth buffer.
In step 120, pixel values are removed from the pipeline based upon the occlusion prediction results. In one exemplary implementation, the pixel values which the prediction indicates will be occluded are removed from the pipeline (e.g., discarded) before occupying intermediate graphics processing pipeline resources (e.g., shading system resources) and performing processing on the predicted occluded pixel values. Processing time is not consumed on pixel values that are not seen. Since pixel values that are occluded and not seen by the user are discarded at the end of the pipeline anyway, they are culled before intermediate processing is performed on them.
In step 130, continued graphics pipeline processing is performed on remaining pixel values. In one embodiment, shading processing is performed upon the remaining pixels. The remaining pixels are also subjected to full occlusion processing in a final occlusion analysis stage (e.g., Z buffer processing) and pixel values that were not culled earlier but are occluded are discarded during the final occlusion analysis stage.
In one embodiment of the present invention, stencil occlusion processing results are also predicted in step 110. For example, pixel values presently in the rasterizing stage of a graphics processing pipeline are examined for potential stencil processing and a prediction regarding the stencil processing is made. In one exemplary implementation, the stencil processing prediction includes a stencil value threshold comparison. If the pixel values presently in the rasterizer stage do not meet the stencil value threshold value, intermediate stencil processing is not performed on the pixel.
It is appreciated that the present invention can be implemented in a variety of embodiments. In one exemplary implementation, the present invention can be utilized in processing systems that support a variety of graphics applications including video games. For example, the present invention can be utilized in graphics rendering processes of a game console, personal computer, personal digital assistant, cell phone or any number of platforms for implementing a video game. It is also appreciated that references to video game application implementations are exemplary and the present invention is not limited to these implementations.
The components of computer system 200 cooperatively operate to provide presentations of graphics images. Communications bus 291 through 297 communicate information. Central processor 201 processes information. Main memory 202 stores information and instructions for the central processor 201. Removable data storage device 204 also stores information and instructions (e.g., functioning as a large information reservoir). Input device 207 provides a mechanism for inputting information and/or for pointing to or highlighting information on display 220. Signal communication port 208 provides a communication interface to exterior devices (e.g., an interface with a network). Display device 209 displays information in accordance with data stored in frame buffer 215. Graphics processor 211 processes graphics commands from central processor 201 and provides the resulting data to graphics buffer 215 for storage and retrieval by display monitor 220.
In one embodiment of the present invention, graphics processor 211 processes information, including graphics information (e.g., information associated with graphics processing method 100). Graphics processor 211 processes the graphics information in a pipeline sequence and performs culling during a rasterization stage of the pipeline. Graphics processor 211 can compare pre-cull values to end of graphics processing pipeline values (e.g., post occlusion determination values) to analyze if pre-cull pixel values should be culled. In one embodiment of the present invention, the cull values are associated with a prediction area. The end of graphics processing pipeline values can also be stored in an on chip register for utilization in future occlusion predictions.
The components of graphics processing pipeline 300 cooperatively operate to perform graphics pipeline operations. Vertex module 310 determines vertices information and forwards the resulting information to rasterizing module 320. Rasterizing module 320 rasterizes pixels including performing culling based upon prediction results. In one exemplary implementation, rasterizing module 320 determines coordinate values (e.g., identifies X, Y and Z coordinate values) of the respectively received graphics pipeline pixel information and forwards graphics pipeline information to shader module 341. Shader module 330 performs pixel shading calculations on the pixel information and forwards the results to raster operation (ROP) component 340. Raster operation (ROP) component 340 performs final processing on pixel values (e.g., occlusion and stencil processing) and writes appropriate values to a frame buffer (not shown). For example, ROP component 340 reads “old” Z values from a Z buffer, compares the “old” Z values to corresponding current Z values and writes the corresponding current Z values to the Z buffer if the current Z values are nearer than the old Z values. ROP component 340 also forwards the farthest Z value in a prediction area to rasterizing module 320. In one embodiment, ROP component 340 also tests stencil values as they are read and/or written to determine if a pixel value update would be rejected. The results for a prediction area are aggregated similar to Z values by recording (e.g., asserting a corresponding bit in a stencil mask) if pixels within the prediction area would cause rejection.
In one exemplary implementation, rasterizing module 320 performs the rasterizing in a coarse rasterization stage and a fine rasterization stage. The occlusion culling can be performed after the coarse stage and before the fine stage and reduce precise pixel rasterization work of the fine stage performed on pixel values that are eventually discarded. This permits pixel values to be discarded at a rate faster than the fine rasterizer operates. Rasterizing module 320 establishes which pre-culled values are associated with pixels within bounds of a prediction area and which vertex of the prediction area is closest or nearest. Rasterizer module 320 compares the closest pre-culled vertex value to a corresponding occlusion value received from raster operation module 340 at the end of graphics processing pipeline 300. If the present closest pre-culled vertex value for the prediction area is behind the corresponding occlusion value for the prediction area received from the end of the graphics processing pipeline, the present group of pixel values associated with the predication area are culled (e.g., discarded) by rasterizer module 330 and not forwarded to the remaining modules for processing. In one embodiment, values received from ROP 340 are “outdated” and the Z values change monotonically closer to the eye. If the Z values change non-monotonically the culling operations are suspended until the start of the next frame.
In one embodiment of the present invention, occlusion prediction information is compressed.
In step 410, a cull value for a pixel culling area is retrieved. The cull value is associated with a pixel culling area (e.g., a 16 by 16 pixel area). In one embodiment, the cull value is retrieved from an on chip memory location (e.g., RAM, cache, etc.). In one exemplary implementation, an initial cull value is established based upon the farthest plane on a depth axis of a viewing volume. In another exemplary implementation, the initial cull value is the farthest depth threshold value of an occluding volume.
In step 420, an end of graphics processing pipeline occlusion value associated with a prediction area within the pixel culling area is received. In one embodiment, an end of graphics processing pipeline occlusion value is received from a raster operation (ROP) component at the end of a graphics processing pipeline. In one exemplary implementation, a plurality of prediction area end of graphics processing pipeline occlusion values are aggregated. The aggregation includes establishing the farthest occlusion value. A mask (e.g., a depth or Z value mask, a stencil mask, etc.) is also produced as part of the aggregation. The mask denotes which prediction areas within the culling area have end of graphics processing pipeline Z values at or nearer than the Z value associated with the mask.
In one embodiment of the present invention, the mask includes occlusion indications for each prediction area within a culling region. For example, a depth or Z value mask indicates which pixels values in a particular culling area are equal to or in front of a Z cull value associated with the mask. In one embodiment, each bit in a mask is associated with a bounded pixel area permitting significant compression. For example, storing a cull Z value and a mask in which each bit represents a 4 by 2 prediction area permits 139 to 1 compression in one exemplary implementation. In one exemplary implementation the cull Z value and mask are stored in a 44 bit word on chip memory word associated with a 16 by 16 pixel cull area or region. The 44 bit memory word comprises a 12 bit cull Z value and 32 bit mask in which each bit represents a 4 by 2 pixel prediction area. In a Z buffer the 16 by 16 cull area Z information is represented by 6,144 bits (16 pixels times 16 pixels times 24 bits).
In step 430, a view volume is segmented into regions based upon offsets from the cull value. In one embodiment, there is a negative offset range value that establishes a near threshold and a positive offset range value that establishes a far threshold value.
In one embodiment of the present invention, software establishes the offset values. For example, the offset values can be established according to the application. In one embodiment of the present invention, the negative cull value and positive cull value are established by drivers.
In step 440, a determination of adjustments to the cull value and associated culling mask is made based upon which of the regions or “buckets” (e.g., 482, 483, 484, or 485) the end of graphics processing pipeline occlusion value is within. If the end of graphics processing pipeline value is nearer than near threshold 472 then the cull value is replaced by the end of graphics processing pipeline occlusion value and the culling mask bits are updated accordingly. For example, culling mask bits corresponding to the prediction area associated with the end of graphics processing pipeline occlusion value are asserted and the remaining culling mask bits are unasserted.
If the end of graphics processing pipeline value is between the near threshold 472 and the cull value 473 then the cull value is left unchanged and the culling mask bits are updated accordingly. For example, the culling mask bits corresponding to the prediction area associated with the end of graphics processing pipeline occlusion value are asserted and the remaining culling mask bits are unchanged. In one exemplary implementation, in which each bit in the mask corresponds to a prediction area, a logical OR operation can be performed on the prediction area bits in the mask and the prediction area bits associated with the end of graphics processing pipeline value to produce updated mask values.
If the end of graphics processing pipeline value is between the cull value 473 and the far threshold 474 then the cull value is replaced by the end of graphics processing pipeline occlusion value and the culling mask bits are updated accordingly. For example, the culling mask bits corresponding to the prediction area associated with the end of graphics processing pipeline occlusion value are asserted and the remaining culling mask bits are unchanged. Again, in one exemplary implementation in which each bit in the mask corresponds to a prediction area, a logical OR operation can be performed on the prediction area bits in the mask and the prediction area bits associated with the end of graphics processing pipeline value to produce updated mask values. If the end of graphics processing pipeline value is farther than threshold 474 no adjustments are made to the cull value or associated mask.
The components of pixel cull value compression system 500 cooperatively operate to compress pixel cull value information. Aggregators 501 through 504 aggregate end of pipe depth value information associated with a bounded pixel area. In one exemplary implementation, aggregators 501 through 504 aggregate end of graphics processing pipeline values received in step 420. Crossbar switch 509 interleaves information associated with a plurality of the bounded pixel areas in a tiled manner. Crossbar switch 509 also balances information associated with the plurality of bounded pixel areas across the plurality of memories 521 through 524. Coalesce components 511A through 514B coalesce depth value information for a plurality of the bounded pixel areas located in adjacent proximity in a presentation plane and replace the updated base cull depth information. In one exemplary implementation, coalesce buffers 511A through 514B segment a view volume into regions based upon offsets from the cull value (e.g., similar to step 430), and determine adjustments to the cull value and cull masks (e.g., similar to step 440).
In one embodiment, coalesce components 511A through 511B include an accumulation register that stores occluder packets as a determination is made as to stores the occlusion information from an end of graphics processing pipeline stage (e.g., raster operation module 340) in a compressed form. In one embodiment of the present invention, the coalesce components 511A through 511B perform segmentation into regions based upon offset from the cull value (e.g., perform step 430). As packets associated with other culling areas begin to arrive and accumulation registers are occupied the current accumulation of information gets displaced and written to on chip memory (e.g., a RAM, cache, etc.). Random access memories 521 through 524 comprise a plurality of memory sub-components and store updated base cull depth information.
Thus, the present invention facilitates optimized utilization of processing resources and conservation of time. Pixels that are eventually occluded are not processed in intermediate stages of a graphics pipeline. By efficiently and effectively utilizing processing resources, end result graphics images processing rate is increased and faster rendering is provided. The present invention also facilitates rate increases because pixels can be culled or rejected faster than they are processed. For example, pixels can be rejected at four times the fastest processing rate and in a frame where 25% of the pixels are culled the overall increase is 23%. In addition, the present invention does not consume additional bandwidth accessing a frame buffer to obtain actual Z or stencil values and updating a compressed representation. The present invention also provides an efficient comparison mechanism without comparing a primitive's (e.g., triangle) full Z range against an occluder Z. The present invention also facilitates conservation of power by not expending energy on intermediate processing for pixels that are eventually discarded before rendering.
The foregoing descriptions of specific embodiments of the present invention have been presented for purposes of illustration and description. They are not intended to be exhaustive or to limit the invention to the precise forms disclosed, and obviously many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and its practical application, to thereby enable others skilled in the art to best utilize the invention and various embodiments with various modifications as are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the Claims appended hereto and their equivalents.
Claims
1. A graphics processing method comprising:
- performing an occlusion prediction in an electronic system based upon an occlusion value received from later in a graphics processing pipeline, wherein said occlusion prediction includes establishing a prediction area and selecting a nearest vertex of said prediction area for prediction analysis,
- removing pixel values from said pipeline based upon results of said occlusion prediction;
- suspending said removing until a next frame if a depth value associated with said occlusion value change non-monotonically; and
- performing continued graphics pipeline processing on remaining pixel values associated with display pixel illumination.
2. A graphics processing method of claim 1 wherein said predicting is performed within a rasterizing stage of a graphic processing pipeline and said occlusion value is received from a raster operation stage.
3. A graphics processing method of claim 1 wherein said removing of said pixels from said pipeline includes pixels values associated with pixels in said prediction area.
4. A graphics processing method of claim 1 wherein said prediction area is a square or rectangle and wherein said nearest vertex is at an intersection of said prediction area side with a primitive side.
5. A graphics processing method of claim 4 wherein said nearest vertex is at a corner of said prediction area.
6. A graphics processing method of claim 4 wherein said nearest vertex is at a vertex of a primitive.
7. A graphics processing method of claim 4 wherein a primitive's depth plane equation is utilized to determine said nearest vertex.
8. A graphics processing method of claim 7 wherein a sign of coefficients of said depth plane equation slope indicates which of said vertices at a corner of said prediction area is a nearest vertex.
9. A graphics processing method of claim 3 wherein estimating in said predicting occlusion results is performed conservatively towards permitting pixel values to proceed to performing said intermediate processing.
10. A graphics processing system comprising:
- a bus for communicating graphics information;
- a memory communicatively coupled to said bus, said memory for storing said graphics information;
- a processor communicatively coupled to said bus, said processor for processing information including pixel values, including performing pixel value occlusion predictions, based upon an occlusion value received from later in a graphics processing pipeline, wherein said occlusion prediction includes establishing a prediction area and selecting a nearest vertex of said prediction area for prediction analysis, and culling associated with said pixel value occlusion predictions is suspended until a next frame if a depth value associated with said pixel value occlusion predictions changes non-monotonically, wherein said pixel values are associated with display pixel illumination.
11. A graphics processing system of claim 10 wherein said processor processes said graphics information in a pipeline sequence and performs said occlusion result prediction during a rasterization stage of said pipeline.
12. A graphics processing system of claim 11 wherein said processor compares pre-cull values to end of graphics processing pipeline occlusion values to analyze if pre-cull pixel values should be culled.
13. A graphics processing system of claim 12 wherein said cull values are associated with a prediction area.
14. A graphics processing system of claim 11 wherein said occlusion values are stored in an on chip register.
15. A computer readable medium with instructions embedded thereon for causing a processor to perform graphics processing, said instructions comprising:
- a rasterizer module for rasterizing pixels and performing an occlusion prediction based upon an occlusion value received from later in a graphics processing pipeline, wherein said occlusion prediction includes establishing a prediction area and selecting a nearest vertex of said prediction area for prediction analysis, and culling associated with said occlusion prediction is suspended until a next frame if a depth value associated with said occlusion prediction changes non-monotonically;
- a shader module for shading remaining pixels; and
- a depth buffer module for discarding occluded pixels and storing visible pixels in a depth buffer, wherein said pixels are associated with display illumination.
16. A non-transitory computer readable medium of claim 15 wherein said rasterizer module performs said rasterizing and occlusion prediction in a coarse rasterization stage and a fine rasterization stage.
17. A non-transitory computer readable medium of claim 15 wherein said rasterizer module establishes which pixels are within bounds of a prediction area and which vertex of said prediction area is nearest.
18. A non-transitory computer readable medium of claim 17 wherein said nearest vertex is at a corner of said prediction area.
19. A non-transitory computer readable medium of claim 15 wherein an occlusion process is performed in which pixels that are not visible in a presentation are discarded.
20. A non-transitory computer readable medium of claim 15 wherein said nearest vertex is at an intersection of said prediction area side with a primitive side.
| 4958305 | September 18, 1990 | Piazza |
| 5043921 | August 27, 1991 | Gonzalez-Lopez et al. |
| 5369741 | November 29, 1994 | Hartog et al. |
| 5583974 | December 10, 1996 | Winner et al. |
| 5596686 | January 21, 1997 | Duluk, Jr. |
| 5694143 | December 2, 1997 | Fielder et al. |
| 5751291 | May 12, 1998 | Olsen et al. |
| 5761400 | June 2, 1998 | Derby et al. |
| 5909595 | June 1, 1999 | Rosenthal et al. |
| 5914721 | June 22, 1999 | Lim |
| 5990904 | November 23, 1999 | Griffin |
| 6052128 | April 18, 2000 | Narayanaswami et al. |
| 6052129 | April 18, 2000 | Fowler et al. |
| 6091428 | July 18, 2000 | Piazza et al. |
| 6094200 | July 25, 2000 | Olsen et al. |
| 6172670 | January 9, 2001 | Oka et al. |
| 6246415 | June 12, 2001 | Grossman et al. |
| 6323860 | November 27, 2001 | Zhu et al. |
| 6373485 | April 16, 2002 | Sowizral et al. |
| 6407741 | June 18, 2002 | Morein et al. |
| 6445390 | September 3, 2002 | Aftosmis et al. |
| 6476807 | November 5, 2002 | Duluk, Jr. et al. |
| 6480205 | November 12, 2002 | Greene et al. |
| 6525726 | February 25, 2003 | Xie et al. |
| 6636212 | October 21, 2003 | Zhu |
| 6636215 | October 21, 2003 | Greene |
| 6646639 | November 11, 2003 | Greene et al. |
| 6677945 | January 13, 2004 | Lapidous et al. |
| 6717576 | April 6, 2004 | Duluk, Jr. et al. |
| 6720964 | April 13, 2004 | Fowler et al. |
| 6862025 | March 1, 2005 | Buehler |
| 6900800 | May 31, 2005 | Baldwin |
| 6924801 | August 2, 2005 | Dorbie |
| 7030877 | April 18, 2006 | Sell |
| 7091971 | August 15, 2006 | Morein |
| 7218317 | May 15, 2007 | Liao et al. |
| 7242400 | July 10, 2007 | Van Hook et al. |
- M. Meiβner, D. Bartz, and R. Günther and W. Straβer, “Visibility Driven Rasterization,” 2001, Computer Graphics Forum, vol. 20 No. 4, p. 283-294.
- NVID-P000794; Douglas A. Voorhies; Occulusion Prediction Compression System and Method; Appl. No.: 10/745,277; Filing Date: Dec. 22, 2003.
- NVID-P000793.CIP.1; Douglas A. Voorhies; Culling Data Selection System and Method; Appl. No.: 11/303,351; Filing Date: Dec. 16, 2005.
- NVID-P00793.CIP.2; Douglas A. Voorhies; Tight Depth Range Occlusion Prediction System and Method; Appl. No.: 11/397,528; Filing Date: Apr. 03, 2006.
- M.J. Aftosmis, “Solution Adaptive Cartesian Grid Methods for Aerodynamic Flows With Complex Geometrics”, Mar. 3, 1997, Lecture notes for 28th Computational Fluid Dynamics Lecture Series, von Kaman Institute for Fluid Dynamics, pp. 88-96.
- Michael D. McCool, Chris Wales, Kevin Moule, “Incremental and Hierarchical Hilbert ORder Edge Equeation Polygon Rasterization”, Aug. 2001, Proceedings of the ACM SIGGRAPH/EUROGRAPHICS workshop on Graphics Hardware, pp. 65-72d.
Type: Grant
Filed: Dec 22, 2003
Date of Patent: Aug 23, 2011
Assignee: NVIDIA Corporation (Santa Clara, CA)
Inventors: Douglas A. Voorhies (Menlo Park, CA), Svetoslav D. Tzvetkov (Irvine, CA)
Primary Examiner: Jason M Repko
Application Number: 10/745,160
International Classification: G06T 15/40 (20060101);
