公开(公告)号：US07830390B2

公开(公告)日：2010-11-09

申请号：US11183848

申请日：2005-07-19

申请人： David Shreiner

发明人： David Shreiner

IPC分类号： G06F15/80 ,  G06F15/16

CPC分类号： G06T15/005 ,  G06T15/80

摘要： A plurality of vertex or fragment processors on a graphics processor perform computations. Each vertex or fragment processor is capable of executing a separate program to compute a specific result. A combiner manages the combination of the results from the respective processors, and produces a final transformed vertex or pixel value. The vertex or fragment processors and the combiner can be programmable to modify their operations. As such, the vertex or fragment processors can operate in a parallel or serial configuration, or both. The combiner manages and resolves the operations of the serial and/or parallel configurations. A synchronization barrier enables the combiner to perform data-dependency analysis to determine the timing and ordering of the respective processors' execution. A transformation module can include one or more programmable vertex processors that transforms three-dimensional geometric data into fragments. The fragment processors subsequently receive and compute a result for each fragment, which is combined using the combiner to produce the final pixel value.

摘要翻译： 图形处理器上的多个顶点或片段处理器执行计算。 每个顶点或片段处理器能够执行单独的程序来计算特定的结果。 组合器管理来自各个处理器的结果的组合，并且产生最终变换的顶点或像素值。 顶点或片段处理器和组合器可以被编程以修改其操作。 因此，顶点或片段处理器可以以并行或串行配置操作，或两者兼而有之。 组合器管理和解决串行和/或并行配置的操作。 同步屏障使得组合器能够执行数据依赖性分析以确定各个处理器执行的时序和顺序。 变换模块可以包括将三维几何数据变换成片段的一个或多个可编程顶点处理器。 片段处理器随后接收并计算每个片段的结果，其使用组合器组合以产生最终像素值。

公开(公告)号：US20070018990A1

公开(公告)日：2007-01-25

申请号：US11183848

申请日：2005-07-19

申请人： David Shreiner

发明人： David Shreiner

IPC分类号： G06F15/80

CPC分类号： G06T15/005 ,  G06T15/80

摘要： A plurality of vertex or fragment processors on a graphics processor perform computations. Each vertex or fragment processor is capable of executing a separate program to compute a specific result. A combiner manages the combination of the results from the respective processors, and produces a final transformed vertex or pixel value. The vertex or fragment processors and the combiner can be programmable to modify their operations. As such, the vertex or fragment processors can operate in a parallel or serial configuration, or both. The combiner manages and resolves the operations of the serial and/or parallel configurations. A synchronization barrier enables the combiner to perform data-dependency analysis to determine the timing and ordering of the respective processors' execution. A transformation module can include one or more programmable vertex processors that transforms three-dimensional geometric data into fragments. The fragment processors subsequently receive and compute a result for each fragment, which is combined using the combiner to produce the final pixel value.

摘要翻译： 图形处理器上的多个顶点或片段处理器执行计算。 每个顶点或片段处理器能够执行单独的程序来计算特定的结果。 组合器管理来自各个处理器的结果的组合，并且产生最终变换的顶点或像素值。 顶点或片段处理器和组合器可以被编程以修改其操作。 因此，顶点或片段处理器可以以并行或串行配置操作，或两者兼而有之。 组合器管理和解决串行和/或并行配置的操作。 同步屏障使得组合器能够执行数据依赖性分析以确定各个处理器执行的时序和顺序。 变换模块可以包括将三维几何数据变换成片段的一个或多个可编程顶点处理器。 片段处理器随后接收并计算每个片段的结果，其使用组合器组合以产生最终像素值。

公开(公告)号：US07716683B2

公开(公告)日：2010-05-11

申请号：US11027752

申请日：2004-12-30

申请人： Hansong Zhang ,  David Shreiner

发明人： Hansong Zhang ,  David Shreiner

IPC分类号： G06F9/44 ,  G09G5/39

CPC分类号： G06F13/102 ,  G06F3/14

摘要： A method and apparatus forward a hardware call from a driver to graphics hardware via a virtual connection. Specifically, the method and apparatus process graphical data in a system having the driver, which produces a hardware call for the controlling the operation of the graphics hardware. As noted above, the method and apparatus first establish the virtual connection between the driver and the graphics hardware. Next, the hardware call is forwarded to the graphics hardware via the virtual connection.

摘要翻译： 一种方法和装置，通过虚拟连接将硬件调用从驱动程序转发到图形硬件。 具体地，该方法和装置处理具有驱动器的系统中的图形数据，其产生用于控制图形硬件的操作的硬件调用。 如上所述，该方法和装置首先建立驱动程序和图形硬件之间的虚拟连接。 接下来，通过虚拟连接将硬件调用转发到图形硬件。

公开(公告)号：US20060146058A1

公开(公告)日：2006-07-06

申请号：US11027752

申请日：2004-12-30

申请人： Hansong Zhang ,  David Shreiner

发明人： Hansong Zhang ,  David Shreiner

IPC分类号： G09G5/39

CPC分类号： G06F13/102 ,  G06F3/14

摘要： A method and apparatus forward a hardware call from a driver to graphics hardware via a virtual connection. Specifically, the method and apparatus process graphical data in a system having the driver, which produces a hardware call for the controlling the operation of the graphics hardware. As noted above, the method and apparatus first establish the virtual connection between the driver and the graphics hardware. Next, the hardware call is forwarded to the graphics hardware via the virtual connection.

摘要翻译： 一种方法和装置，通过虚拟连接将硬件调用从驱动程序转发到图形硬件。 具体地，该方法和装置处理具有驱动器的系统中的图形数据，其产生用于控制图形硬件的操作的硬件调用。 如上所述，该方法和装置首先建立驱动程序和图形硬件之间的虚拟连接。 接下来，通过虚拟连接将硬件调用转发到图形硬件。

公开(公告)号：US20060093044A1

公开(公告)日：2006-05-04

申请号：US11209765

申请日：2005-08-24

申请人： Brad Grantham ,  David Shreiner ,  Alan Commike

发明人： Brad Grantham ,  David Shreiner ,  Alan Commike

IPC分类号： H04N7/04

CPC分类号： G06F9/3851 ,  G06F3/1446 ,  G06T15/005

摘要： A system and method for distributing data (e.g., imaging data such as pixels, or 3D graphics data such as points, lines, or polygons) from a single or a small number of data sources to a plurality of graphical processing units (graphics processors) for processing and display is presented. The system and method provide a pipelined and multithreaded approach that prioritizes movement of the data through a high-speed multiprocessor system (or a high-speed system of networked computers), according to the system topology. Multiple threads running on multiple processors in shared memory move the data from a storage device (e.g., a disk array), through the high-speed multiprocessor system, to graphics processor memory for display and optional processing through fragment programming. The data can also be moved in the reverse direction, back through the high-speed multiprocessor system, for storage on the disk array.

摘要翻译： 一种用于将数据（例如，诸如像素的成像数据或诸如点，线或多边形的3D图形数据）从单个或少数数据源分发到多个图形处理单元（图形处理器）的系统和方法， 用于处理和显示。 该系统和方法提供了一种流水线和多线程方法，根据系统拓扑，通过高速多处理器系统（或联网计算机的高速系统）来优先考虑数据的移动。 在共享存储器中的多个处理器上运行的多个线程将数据从存储设备（例如，磁盘阵列）通过高速多处理器系统移动到图形处理器存储器，用于通过片段编程进行显示和可选处理。 数据也可以反向移动，通过高速多处理器系统，以便存储在磁盘阵列上。

公开(公告)号：US08194083B2

公开(公告)日：2012-06-05

申请号：US12941423

申请日：2010-11-08

申请人： David Shreiner

发明人： David Shreiner

IPC分类号： G06F15/80 ,  G06F15/16

CPC分类号： G06T15/005 ,  G06T15/80

摘要： A plurality of vertex or fragment processors on a graphics processor perform computations. Each vertex or fragment processor is capable of executing a separate program to compute a specific result. A combiner manages the combination of the results from the respective processors, and produces a final transformed vertex or pixel value. The vertex or fragment processors and the combiner can be programmable to modify their operations. As such, the vertex or fragment processors can operate in a parallel or serial configuration, or both. The combiner manages and resolves the operations of the serial and/or parallel configurations. A synchronization barrier enables the combiner to perform data-dependency analysis to determine the timing and ordering of the respective processors' execution. A transformation module can include one or more programmable vertex processors that transforms three-dimensional geometric data into fragments. The fragment processors subsequently receive and compute a result for each fragment, which is combined using the combiner to produce the final pixel value.

摘要翻译： 图形处理器上的多个顶点或片段处理器执行计算。 每个顶点或片段处理器能够执行单独的程序来计算特定的结果。 组合器管理来自各个处理器的结果的组合，并且产生最终变换的顶点或像素值。 顶点或片段处理器和组合器可以被编程以修改其操作。 因此，顶点或片段处理器可以以并行或串行配置操作，或两者兼而有之。 组合器管理和解决串行和/或并行配置的操作。 同步屏障使得组合器能够执行数据依赖性分析以确定各个处理器执行的时序和顺序。 变换模块可以包括将三维几何数据变换成片段的一个或多个可编程顶点处理器。 片段处理器随后接收并计算每个片段的结果，其使用组合器组合以产生最终像素值。

公开(公告)号：US20110050697A1

公开(公告)日：2011-03-03

申请号：US12941423

申请日：2010-11-08

申请人： David SHREINER

发明人： David SHREINER

IPC分类号： G06T15/80 ,  G06F15/80

CPC分类号： G06T15/005 ,  G06T15/80

摘要： A plurality of vertex or fragment processors on a graphics processor perform computations. Each vertex or fragment processor is capable of executing a separate program to compute a specific result. A combiner manages the combination of the results from the respective processors, and produces a final transformed vertex or pixel value. The vertex or fragment processors and the combiner can be programmable to modify their operations. As such, the vertex or fragment processors can operate in a parallel or serial configuration, or both. The combiner manages and resolves the operations of the serial and/or parallel configurations. A synchronization barrier enables the combiner to perform data-dependency analysis to determine the timing and ordering of the respective processors' execution. A transformation module can include one or more programmable vertex processors that transforms three-dimensional geometric data into fragments. The fragment processors subsequently receive and compute a result for each fragment, which is combined using the combiner to produce the final pixel value.

摘要翻译： 图形处理器上的多个顶点或片段处理器执行计算。 每个顶点或片段处理器能够执行单独的程序来计算特定的结果。 组合器管理来自各个处理器的结果的组合，并且产生最终变换的顶点或像素值。 顶点或片段处理器和组合器可以被编程以修改其操作。 因此，顶点或片段处理器可以以并行或串行配置操作，或两者兼而有之。 组合器管理和解决串行和/或并行配置的操作。 同步屏障使得组合器能够执行数据依赖性分析以确定各个处理器执行的时序和顺序。 变换模块可以包括将三维几何数据变换成片段的一个或多个可编程顶点处理器。 片段处理器随后接收并计算每个片段的结果，其使用组合器组合以产生最终像素值。

公开(公告)号：US07460126B2

公开(公告)日：2008-12-02

申请号：US11209765

申请日：2005-08-24

申请人： Brad Grantham ,  David Shreiner ,  Alan Commike

发明人： Brad Grantham ,  David Shreiner ,  Alan Commike

IPC分类号： G06F15/80 ,  G06F15/16 ,  G06F15/167 ,  G06T1/20

CPC分类号： G06F9/3851 ,  G06F3/1446 ,  G06T15/005

摘要： A system and method for distributing data (e.g., imaging data such as pixels, or 3D graphics data such as points, lines, or polygons) from a single or a small number of data sources to a plurality of graphical processing units (graphics processors) for processing and display is presented. The system and method provide a pipelined and multithreaded approach that prioritizes movement of the data through a high-speed multiprocessor system (or a high-speed system of networked computers), according to the system topology. Multiple threads running on multiple processors in shared memory move the data from a storage device (e.g., a disk array), through the high-speed multiprocessor system, to graphics processor memory for display and optional processing through fragment programming. The data can also be moved in the reverse direction, back through the high-speed multiprocessor system, for storage on the disk array.

摘要翻译： 一种用于将数据（例如，诸如像素的成像数据或诸如点，线或多边形的3D图形数据）从单个或少数数据源分发到多个图形处理单元（图形处理器）的系统和方法， 用于处理和显示。 该系统和方法提供了一种流水线和多线程方法，根据系统拓扑，通过高速多处理器系统（或联网计算机的高速系统）来优先考虑数据的移动。 在共享存储器中的多个处理器上运行的多个线程将数据从存储设备（例如，磁盘阵列）通过高速多处理器系统移动到图形处理器存储器，用于通过片段编程进行显示和可选处理。 数据也可以反向移动，通过高速多处理器系统，以便存储在磁盘阵列上。