Culling data selection system and method
Patent 7619625 Issued on November 17, 2009. Estimated Expiration Date: 
December 16, 2025. Estimated Expiration Date is calculated based on simple USPTO term provisions. It does not account for terminal disclaimers, term adjustments, failure to pay maintenance fees, or other factors which might affect the term of a patent.
December 16, 2025. Estimated Expiration Date is calculated based on simple USPTO term provisions. It does not account for terminal disclaimers, term adjustments, failure to pay maintenance fees, or other factors which might affect the term of a patent.Patent ReferencesPolygon edge clipping High speed Z-buffer control Method for pre-clipping a line lying within a clipping rectangular region which is a subset of a region of a display screen Computer graphics system having high performance multiple layer Z-buffer Method and apparatus for simultaneous parallel query graphics rendering Z-coordinate buffer System and method for accelerated occlusion culling Method and system for increasing the speed of a Z-buffer process Method of controlling I/O routing by setting connecting context for utilizing I/O processing elements within a computer system to produce multimedia effects Visibility calculations for 3D computer graphics Method and apparatus for clipping convex polygons on single instruction multiple data computers InventorAssigneeApplicationNo. 11303351 filed on 12/16/2005US Classes:345/422Z buffer (depth buffer)ExaminersPrimary: Chauhan, UlkaAssistant: Repko, Jason M International ClassG06T 15/40DescriptionFIELD OF THE INVENTIONThe present invention relates to the field of graphics processing. BACKGROUND OF THE INVENTION Electronic 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 digitalcomputers, 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 whichtraditionally 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 graphicsystems 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 parametervalues 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 arelatively 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 thatare 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 pixelvalues. 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 userexperience 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 increasingdemands 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 andcomplex 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 theincreased 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 thegraphics presentation. SUMMARY An culling data selection 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 pixels are eventually occludedbefore intermediate processing stages are performed on the pixels. Culling information utilized to predict which pixel are occluded is selected and compressed in accordance with embodiments of the present invention. In one embodiment, cull data for apixel culling area is retrieved and an end of pipe depth occlusion data associated with a prediction area within the pixel culling area is received. A selection metric for analyzing adjustments to cull data is established and a cull data adjustmentdecision is made based upon the selection metric. In one exemplary implementation the possible occlusion volumes associated with "old" culling data, "new" culling data (e.g., based upon an occlusion value received from a stage later in a graphicsprocessing pipeline) and "merged" culling data are determined. The culling data associated with the greatest occlusion volume is selected for culling operations. DESCRIPTION OF THE DRAWINGS 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 beunderstood as not being drawn to scale except if specifically noted. FIG. 1 is a flow chart of a graphics processing method in accordance with one embodiment of the present invention. FIG. 2 is a block diagram of a computer system in accordance with one embodiment of the present invention. FIG. 3 is a block diagram of a graphics pipeline in accordance with one embodiment of the present invention. FIG. 4 is a flow chart of an exemplary pixel cull data method in accordance with one embodiment of the present invention. FIG. 5 is a block diagram of a pixel cull value compression system in accordance with one embodiment of the present invention. DETAILED DESCRIPTION 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 beunderstood 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 asdefined 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 toone 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 obscureaspects 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 andrepresentations 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 aself-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 quantumsignals 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 fromthe 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 computersystem, 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 processingdevices 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 quantitieswithin other components. In one embodiment of the present invention, culling data is selected based upon an occlusion assessment metric and utilized in an occlusion prediction process. The occlusion assessment metric provides an indication of the probability that animage is "occluded" by the culling data and a pixel associated with the image can be eliminated early in a pipeline. The greater the occlusion assessment metric the greater the probability that the drawing is occluded by the culling data. In oneembodiment, the occlusion assessment metric is corresponds to an occlusion volume. In one exemplary implementation, the occlusion volume is the "distance to a far plane" (e.g., the Z difference between a Z limit and the far Z plane) multiplied by an XYindication (e.g., a number of bits in a mask associated with the culling data). In one embodiment, the occlusion volume is a portion of a viewing volume. A 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". FIG. 1 is a flow chart of occlusion prediction graphics processing method 100 in accordance with one embodiment of the present invention. Occlusion prediction graphics processing method 100 facilitates efficient and effective utilization ofprocessing resources. In one embodiment, occlusion prediction graphics processing method 100 is utilized to predict which pixel values are eventually occluded before intermediate processing stages are performed on the pixels. Pixel values are removedor discarded from the pipeline based upon occlusion prediction results and do not unnecessarily occupy processing resources. Remaining pixels are processed through the graphics processing pipeline. In step 110, an occlusion prediction is performed based upon culling data selected in accordance with results of a culling data selection process which includes analysis of occlusion values received from a stage later in a graphics processingpipeline. The culling data assessment results include an occlusion assessment metric that provides an indication of the probability that an image is "occluded" by the culling data. In one embodiment, the culling data selection process includesdetermining possible occlusion volumes. The possible occlusion volumes associated with "old" culling data, "new" culling data and "merged" culling data are determined. The new culling data is based upon an occlusion value received from a stage later ina graphics processing pipeline. The culling data associated with the greatest occlusion volume is selected for culling operations. Receiving the "new" 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., froma 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., toa 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 arasterizing stage and a new occlusion value is 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 analysis indicates that the pixels in the prediction area currently being rasterized are behind the culling data, the prediction is the pixels currently being rasterized shouldbe 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 permitsthe 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 orrectangular 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 orclosest 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 ismade 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. Sincepixel 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 one embodiment, a pixel associated with X,Y position corresponding to theculling data is removed from the pipeline if the pixel Z value is behind culling data Z values. 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 afinal 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 stencilprocessing 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 meetthe 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 includingvideo 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 alsoappreciated that references to video game application implementations are exemplary and the present invention is not limited to these implementations. FIG. 2 is a block diagram of a computer system 200, one embodiment of a computer system upon which embodiments of the present invention can be implemented. Computer system 200 includes central processor unit 201, main memory 202 (e.g., randomaccess memory), chip set 203 with north bridge 209 and south bridge 205, removable data storage device 204, input device 207, signal communications port 208, and graphics subsystem 210 which is coupled to display 220. Computer system 200 includesseveral busses for communicatively coupling the components of computer system 200. Communication bus 291 (e.g., a front side bus) couples north bridge 209 of chipset 203 to central processor unit 201. Communication bus 292 (e.g., a main memory bus)couples north bridge 209 of chipset 203 to main memory 202. Communication bus 293 couples north bridge of chipset 203 to graphic subsystem 210. Communication buses 294-297 (e.g., a PCI bus) couple south bridge 205 of chip set 203 to removable datastorage device 204, input device 207, and signal communications port 208 respectively. Graphics subsystem 210 includes graphics processor 211 and graphics buffer 215. 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 storesinformation 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 informationand/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 datastored 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 graphicsinformation 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 analyzeif 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 infuture occlusion predictions. FIG. 3 is a block diagram of graphics processing pipeline 300 in accordance with one embodiment of the present invention. Graphics processing pipeline 300 (e.g., a pixel processing pipeline) comprises pipeline vertex module 310, rasterizingmodule 320, shader module 330, and raster operation module 340. Vertex module 310, rasterizing module 320, shader module 330, and raster operation module 340 are serially communicatively coupled to one another. Raster operation module 340 is alsocommunicatively coupled to rasterizing module 320. In one embodiment, graphics processing pipeline 300 is similar to a pipeline process utilized in computer system 200. 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. Rasterizingmodule 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 receivedgraphics 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. Rasteroperation (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, comparesthe "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 torasterizing module 320. In one embodiment, ROP component 340 also performs stencil operations and 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 areaggregated 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 reduceprecise 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 whichpre-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 receivedfrom 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 thegraphics 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 fromROP 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, values from ROP 304 are selected andstored before a comparison is performed. In one embodiment of the present invention, occlusion prediction information is compressed. FIG. 4 is a flow chart of pixel cull value compression selection method 400 in accordance with one embodiment of the present invention. In oneembodiment, the cull Z value and an associated mask are stored in an on chip memory (e.g., random access memory, cache, etc.). The mask denotes fully occluded pixels at the corresponding cull value or in front of the cull value. The mask is stored in amemory word with the cull value. In one embodiment of the present invention, pixel cull value compression method 400 is used to compress occlusion prediction values utilized in occlusion prediction graphics processing method 100. In step 410, cull data associated with a pixel culling area (e.g., a 16 by 16 pixel area, 8×8 pixel area, 4×8 pixel area, etc.) is retrieved. In one embodiment, the cull data is retrieved from an on chip memory location (e.g., RAM,cache, etc.). In one exemplary implementation, initial cull data 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 representable depth thresholdvalue of an occluding volume. In step 420, an end of graphics processing pipeline occlusion data associated with a prediction area within the pixel culling area is received. In one embodiment, an end of graphics processing pipeline occlusion data is received from a rasteroperation (ROP) component at the end of a graphics processing pipeline. In one embodiment, the culling data and end of graphics processing pipeline occlusion data include aggregated pixel information. The aggregation includes establishing the farthestocclusion 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 nearerthan 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 orin 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 2prediction 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 bitmemory 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 oneembodiment, a mask denotes areas with Z that is nearer than the associated Z value and the mask bit is asserted if the corresponding area is fully covered by the updates from the ROP. In step 430, selection metrics are established for analyzing adjustments to the cull data. In one embodiment, establishing selection metrics includes determining the volume behind the current cull value, a present retrieved end of graphicsprocessing pipeline occlusion value and a merged configuration value respectively. The volume is determined by multiplying the respective XY indications by the respective Z depths. In one embodiment, the cull data and occlusion data include arespective single farthest Z limit value and a respective mask denoting which pixel blocks are known to be one side of the farthest Z limit. In one exemplary implementation, the respective occlusion volume is the respective "distance to a far plane"(e.g., the Z difference between a Z limit and the far Z plane) multiplied by the respective XY information (e.g., respective XY area, number of asserted bits in a respective mask, etc). In one embodiment, the selection metrics are weighted to compensate for crowding. For example, since Z values tend to be crowded at the far plane when doing perspective from the eye. The Z values can be weighted more or biased to account forcloseness to the far plane since the first little distance from the far plane is "more" valuable in terms of occluded pixels than a corresponding distance closer to the eye. In one embodiment, weighting is performed in accordance with a currentapplication. For example, the pipeline settings can be analyzed to determine if an application is a perspective or orthographic application and weighting is turned on for the perspective application. In one exemplary implementation, the Z values arerepresented in floating points and just looking at the upper bits of the exponent tends to be a logarithmic weighting in the correct direction. In step 440, a cull data adjustment decision is made based upon the selection metrics. The cull data is updated in accordance with the decision and stored in the random access memory and/or an on chip memory. In one embodiment, a culladjustment decision is based upon three possible options. The current cull data (e.g., current cull Z value and corresponding mask) is maintained or unchanged if the selection metric for the current cull value is highest (e.g., the volume is thegreatest). The current cull data is changed to the present retrieved end of graphics pipeline occlusion value and corresponding mask if the selection metric for the present retrieved occlusion value is highest. Current cull data is changed to mergeddata if the selection metric for the merged value is highest. The merged data includes the farthest Z limit value of the current cull data and the present retrieved end of graphics pipeline occlusion data. The merged data also includes logically OR-ingthe current cull mask bits and present retrieved end of graphics pipeline occlusion data. In one embodiment, additional multiplicative or additive adjustments can be made to selection metrics to compensate for a variety of conditions. For example, the merged data can be given additional weight in the case of a "tie" (e.g., the mergevolume and another volume are the same or close). In one exemplary implementation, the merge data is selected if there is a non-monotonic operation (e.g., a clear operation in which Z value is pushed back) as opposed to a monotonic operation (e.g., onethat gets monotonically closer to the eye). FIG. 5 is a block diagram of pixel cull value compression system 500 in accordance with one embodiment of the present invention. Pixel cull value compression system 500 includes aggregators 501 through 504, crossbar switch 509, coalescecomponents 511A through 514B, and random access memories (RAMs) 521 through 524. Aggregators 501 through 504 are coupled to crossbar switch 509 which is coupled to coalesce components 511A through 514B. Coalesce components 511A and B are coupled to RAM521. Coalesce components 512A and B are coupled to RAM 522. Coalesce components 513 A and B are coupled to RAM 523. Coalesce components 514A and B are coupled to RAM 524. 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 oneexemplary 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 boundedpixel areas located in adjacent proximity in a presentation plane and replace the updated base cull depth information. In one embodiment, the base cull depth value is selected based upon present invention selection metrics. In one exemplaryimplementation, establishing selection metrics includes determining the volume behind the current cull value, a present retrieved end of graphics processing pipeline occlusion value and a merged configuration value volume. After establishing theselection metrics, the mask and depth value corresponding to the largest volume is chosen as the cull depth value. For the "old" the old Z and mask are selected, for the "new" the new Z and mask are selected, and for the "merged" the mask is the OR ofthe "old" and "new" mask and the farther of the old and new Z values. 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), anddetermine 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 to store the occlusion information from an end of graphics processing pipeline stage (e.g., rasteroperation 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 associatedwith 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 pluralityof memory sub-components and store updated base cull depth information. Thus, the present invention facilitates optimized utilization of processing resources and conservation of time. Intermediate graphics pipeline stage processing of pixels that are eventually occluded is reduced. By efficiently and effectivelyutilizing 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. Forexample, 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 toobtain 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 inventionalso 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, andobviously 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 skilledin 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. Other References
|
