Computer graphics processing and selective visual display system – Computer graphics display memory system – For storing condition code – flag or status
Reexamination Certificate
2001-08-01
2004-09-14
Tung, Kee M. (Department: 2676)
Computer graphics processing and selective visual display system
Computer graphics display memory system
For storing condition code, flag or status
C345S614000, C345S582000
Reexamination Certificate
active
06791558
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to a three-dimensional rendering processor for processing three-dimensional (3D) graphics, and more particularly, to a method and apparatus for processing pixel rasterization in a 3D rendering processor.
2. Background of the Related Art
Currently, a high-performance 3D rendering processor supports a texture mapping, bump mapping, and environment mapping in addition to an interpolation and depth test (or z test) which are the basic function of the processor.
Especially, the texture mapping is the most important barometer when the performance of the 3D rendering processor is measured, and the importance of the bump mapping and environment mapping has been recently highlighted.
Meanwhile, in order to support the texture mapping, a large amount of memory is required as well as a large-scaled hardware, and this is the most important factor in a memory bandwidth problem that is the biggest bottleneck to the design of the rendering processor.
The most important thing in the design of the rendering processor is how to process the texture mapping effectively.
A pixel rasterization processing section of a general 3D graphic rendering processor performs the depth test after performing the texture mapping.
This is called a pre-texturing method, and the biggest defect of this method is that the texture mapping is performed even with respect to the present image that is superimposed and is not shown due to the previously processed image.
Thus, the pre-texturing method should use an unnecessary memory bandwidth, and this causes the whole performance to deteriorate.
Generally, the 3D graphic image has a plurality of depth complexity, and in case that the depth complexity is 3, about 40% of images are thrown away during the depth test.
Here, the depth complexity represents the average number of fragment images corresponding to one pixel position, and the depth test is for determining the depth value of the present fragment image by comparing the depth value of the previously processed fragment image with that of the fragment image to be presently processed.
In this case, if the depth test is performed after the texture mapping, about 40% of the memory bandwidth required for the whole texture mapping is wasted, and the whole performance deteriorates with the greatly increased power consumption caused by the unnecessary performing of hardware.
In order to solve these problems, a post-texturing method has been proposed.
This method requires a wider fragment image queue in comparison to the pre-texturing method.
Here, the fragment image queue means a FIFO (first-in-first-out) queue for processing a pipeline of the pixel rasterization of the rendering processor.
Also, the post-texturing method includes many processing steps between depth reading operation and writing operation to process a transparent texture image.
Due to this, two or more pixels may share the same memory address between the depth reading operation and writing operation.
Accordingly, the post-texturing method may cause a consistency problem.
In order to detect such a consistency problem, an associative logic is required, but selection of this associative logic imposes a great burden on hardware.
Also, except the rendering processor having a built-in frame memory, most rendering processors use a pixel cache memory having depth information and color information.
However, the hit ratio of the pixel cache memory greatly deteriorates in comparison to that of the microprocessor.
SUMMARY OF THE INVENTION
Accordingly, the present invention is directed to a method and apparatus for processing pixel rasterization in a 3D rendering processor that substantially obviates one or more problems due to limitations and disadvantages of the related art.
An object of the present invention is to provide a method and apparatus for processing pixel rasterization in a 3D rendering processor that can remove the unnecessary performing of the texture mapping.
Another object of the present invention is to provide a method and apparatus for processing pixel rasterization in a 3D rendering processor that can simply and easily detect the consistency problem.
Still another object of the present invention is to provide a method and apparatus for processing pixel rasterization in a 3D rendering processor that can greatly reduce the stop of performing the pipeline.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
To achieve these objects and other advantages and in accordance with the purpose of the invention, as embodied and broadly described herein, a method of processing pixel rasterization in a 3D rendering processor includes the steps of receiving an input of present fragment image information from an outside, primarily checking existence
onexistence of previous fragment image information corresponding to an address of the inputted present fragment image information, immediately texture-mapping the present fragment image information if the previous fragment image information does not exist as a result of the primary checking, while texture-mapping the present fragment image information after determining a depth value of the present fragment image information if the previous fragment image information exists, determining a transparency value of the texture-mapped present fragment image information, secondarily checking existence
onexistence of the previous fragment image information corresponding to the address of the present fragment image information, determining the depth value of the present fragment image information if the previous fragment image information exists as a result of the secondary checking, storing the determined depth value of the present fragment image information, and determining and storing a color value of the present fragment image information.
Here, the method further includes the step of stopping the input of the present fragment image information and repeatedly checking whether the previous fragment image information corresponding to the address of the present fragment image information exists if the previous fragment image information does not exist as the result of the secondary checking.
Meanwhile, the step of primarily checking the existence
onexistence of the previous fragment image information corresponding to the address of the present fragment image information includes the steps of checking whether the address of the present fragment image information coincides with the address of the previous fragment image information referring to a tag of the address of the present fragment image information by comparing the addresses, requesting the previous fragment image information corresponding to the address of the present fragment image information if the addresses do not coincide with each other, receiving the requested previous fragment image information, and reading the depth value of the received previous fragment image information.
Here, the primary checking step further includes the step of comparing the depth value of the previous fragment image information with that of the present fragment image information if the addresses coincide with each other.
Meanwhile, the step of secondarily checking the existence
onexistence of the previous fragment image information corresponding to the address of the present fragment image information includes the steps of checking whether the address of the present fragment image information coincides with the address of the previous fragment image information referring to a tag of the address of the present fragment image information by comparing the addresses, requestin
Han Tack Don
Kim Il San
Lee Kil Whan
Park Woo Chan
Yang Sung Bong
Sheridan & Ross P.C.
Tung Kee M.
Yonsei University
LandOfFree
Method and apparatus for processing pixel rasterization in... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for processing pixel rasterization in..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for processing pixel rasterization in... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3196848