Abstract: A method of dynamically adjusting a multiple stage pipeline to execute one of a set of instructions, wherein each stage has a latency and performs a selected data operation. An instruction to be executed is received and a number of stages of the pipeline is selected to execute the instruction as needed to perform a corresponding data operation. Unnecessary stages are bypassed to a reduced latency and the instruction is executed with the selected stages.

Abstract: A software graphics engine includes a selectable mode filter for improved texture mapping. An x, y pixel coordinate is mapped into a u, v texture map. The mapped u, v coordinate includes integer and fractional portions. Depending on the location of the coordinate relative to the four nearest texels, which are represented as integers, one of several texture mapping schemes are used to either select or calculate the texel value to be used to render the pixel at the x, y screen location. The three texture mapping schemes include point sampling in which the nearest texel from the texture map is selected, two-texel averaging in which the closest two texels are combined in a weighted average, and four-texel averaging in which the closest four texels are combined in a weighted average. By providing a selectable filter than can perform point sampling or two or four-texel averaging, the speed benefit of point sampling can be approached as well as the superior quality of two- and four-texel averaging.

Abstract: A graphics system including a selectable mode filter for improved texture mapping. An x, y pixel coordinate is mapped into a u, v texture map. The mapped u, v coordinate includes integer and fractional portions. Depending on the location of the coordinate relative to the four nearest texels, which are represented as integers, one of several texture mapping schemes are used to either select or calculate the texel value to be used to render the pixel at the x, y screen location. The three texture mapping schemes include point sampling in which the nearest texel from the texture map is selected, two-texel averaging in which the closest two texels are combined in a weighted average, and four-texel averaging in which the closest four texels are combined in a weighted average. By providing a selectable filter than can perform point sampling or two or four-texel averaging, the speed benefit of point sampling can be approached as well as the superior quality of two and four-texel averaging.

Abstract: An integrated memory controller (IMC) which incorporates novel memory, graphics, and audio processing capabilities in a single logical unit. The IMC includes numerous significant advances which provide greatly increased performance over prior art systems. The integrated memory controller (IMC) includes one or more symmetric memory ports for connecting to system memory. The IMC also includes video outputs, preferably RGB (red, green, blue) outputs as well as horizontal and vertical synchronization signal outputs, to directly drive the video display monitor. The IMC transfers data between the system bus and system memory and also transfers data between the system memory and the video display output, thereby eliminating the need for a separate graphics subsystem. The IMC also improves overall system performance and response using main system memory for graphical information and storage.

Abstract: The present invention includes an integrated resize engine and color compare logic for performing a resize bit block transfer (BitBLT) and a transparency BitBLT in a single operation. A source array of pixels is stretched and/or shrunk based upon control signals. The resized pixel values include red, green, and blue color values which are compared with predetermined color range values stored in register pairs. Preferably a register pair is provided for each color. A set of comparators is provided for each color to compare the register values with the color pixel values and to produce an output signal (IN RANGE) indicating if the color pixel value is within the range established by the register values. Each of the in range signals is provided to multiplex logic which generates a transparency enable (TE) output signal based upon the value of the IN RANGE signals and the value of a SELECT input signal.

Abstract: A dynamic pipeline for a processor, including multiple latch stages for providing data to corresponding operation elements and multiplexers with associated control logic for bypassing one or more latch stages and operation elements to execute simpler instructions. For a graphics processor, multiplexers select input pixel values and alpha blending values from either internal or external sources. The pixel values are processed through an arithmetic and logic unit for performing logic operations with other pixel values or with offset scaler values. The alpha values are inverted for performing alpha blending functions. The pixel and alpha values are then provided to a first set of latches for providing latched data to the inputs of a multiplier. The output of the multiplier and another offset scalar value are provided to a second set of latches for providing latched data to an arithmetic element.

Abstract: A processor for executing display list command packets in processor or coprocessor mode of execution. The processor dynamically switches between the two modes based on the commands or interrupts received. Each display list packet includes a plurality of commands associated with a particular function, where each command includes a field for identifying the number of parameters associated with the command, if any. The parameters immediately follow the instruction in the instruction stream in a sequential format, eliminating address dependency. Each command preferably conforms to the same format regardless of location and mode of execution, so that the software and driver is simplified by not having to generate different code for different locations and modes. Thus, a host CPU executing an application program decides whether certain commands and command packets reside in system memory or within a local memory associated with the processor.

Abstract: Interpolation method and apparatus for rendering polygons into a pixel grid. A software driver receives vertices of each polygon and identifies a main slope traversing the vertical extent of the polygon, where the other sides are opposite slopes. The software driver determines initial and incremental width values for width interpolation, rather than edge walking the opposite slopes. The interpolator logic includes a width counter for loading the width value, so that each orthogonal scan line is complete when the width counter reaches terminal count rather than comparing each pixel with the opposite slope. An interpolation procedure thus allows randomly-oriented triangles and other polygons having a main slope, up to two opposite slopes and up to three orthogonal sides to be drawn in a single command.

Abstract: A processor having two separate and relatively independent memory controllers to achieve a dual interface architecture. A first memory controller is coupled to the host interface for retrieving data and instructions and a second memory controller is coupled to an independent local bus for interfacing with a frame buffer memory. A depth buffer may also be coupled to the local bus if desired. Address multiplexor logic is preferably included to allow either memory controller to address either external bus. Multiplexor and buffer logic is also preferably included to allow data transfer in either direction. Preferably, the processor is a graphics processor and both memory controllers are programmable for different addressing formats, such as linear and X/Y in the preferred embodiment. In this manner, data is transferred from host to local memories, and vice versa, in any desired format without delays due to memory controller reconfiguration.