Microsoft Technology Licensing, LLC

Silicon Graphics, Incorporated

Thomas M Olano

Mark S Peercy

John Milligan Airey

A method and system are provided for executing SIMD instructions using graphics technology. A SIMD instruction is received and interpreted. The specific data needed for the SIMD instruction is identified. Texel addresses where the specific data are stored are recalled and frame buffer pixels to be used to support the SIMD instruction are selected. In an alternative embodiment, these texel addresses are stored in frame buffer pixel channels such that the pixel containing a particular address will be the pixel to hold the data stored at that address for the SIMD operation.

Method and system for executing SIMD instructions using graphics technology

One aspect of the invention is a method for accelerating noise. The method includes the step of creating a plurality noise texture values (n , . . . , n ) each corresponding to one of a plurality of image pixel values (x , . . . , x ). The method also includes invoking at least a portion of a graphics pipeline ( ) to blend each of the plurality of image pixel values (x , . . . , x ) with one of the plurality of noise texture values (n , . . . , n ).

Method and system for accelerating noise

Method, system, and computer program product for implementing derivative operators at an interactive rate in computer graphics systems. In an embodiment, a data structure representing the operation of one or more computer program statements is received by a compiler and transformed into a tree data structure. Nodes of the tree data structure that determine derivative information are marked, and a transform rule is applied to the marked nodes to transform the tree data structure into a data structure that can be implemented with graphics hardware interface program statements. In one embodiment of the invention, the compiler transforms shading language statements into graphics application programming interface statements that can be implemented with multiple passes through a graphics pipeline. It is a feature of the present invention that it can be implemented on a wide range of computer graphics systems.

Method, system, and computer program product for implementing derivative operators with graphics hardware

One aspect of the invention is a method for implementing a graphics control construct. The method includes the step of assigning by a graphics system interface ( ) a value to at least one bit in a stencil buffer (S) in response to a first condition of a first conditional clause in an application program ( ). The method also includes the steps of selecting at least a first portion of image data in response to the value, and processing the selected first portion of image data in a frame buffer ( ) of a graphics pipeline ( ).

Method and system for implementing graphics control constructs

A level of detail shading function is produced and stored in a computer readable memory. The level of detail shading function is produced by receiving a shading function, identifying in the shading function at least one candidate block of code for simplification, and generating, for each candidate block of code, at least one simplified block of code that can be substituted for the candidate block of code during image rendering. Candidate blocks of code and simplified blocks of code according to the invention are associated with at least one input parameter and assembled to form the level of detail shading function. During the rendering of an object, input parameters are provided to the level of detail shading function. These input parameters are associated with one or more blocks of code in the level of detail shading function. The input parameters specify how the object is to be shaded using the level of detail shading function.

Growing community of inventors

San Francisco, CA, United States of America

Thomas M Olano