专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

81. 发明授权

US09183014B2 Enabling virtual calls in a SIMD environment 有权
标题翻译：在SIMD环境中启用虚拟呼叫
公开(公告)号：US09183014B2
公开(公告)日：2015-11-10
申请号：US13028574
申请日：2011-02-16
申请人： Wei-Yu Chen , Guei-Yuan Lueh , Subramaniam Maiyuran
发明人： Wei-Yu Chen , Guei-Yuan Lueh , Subramaniam Maiyuran
IPC分类号： G06F15/80 , G06F9/455 , G06F9/30 , G06F9/44
CPC分类号： G06F9/547 , G06F9/30061 , G06F9/449 , G06F9/455 , G06F15/8007
摘要： Systems and methods of enabling virtual calls in a single instruction multiple data (SIMD) environment may involve detecting a virtual call of a function and using a single dispatch of the function to invoke the virtual call for two or more channels of the virtual call. In one example, it is determined that the two or more channels share a common target address and a single dispatch of the function is conducted with respect to the common target address. The process may be iterated for additional channels of the virtual call that share a common target address.
摘要翻译：在单个指令多数据（SIMD）环境中启用虚拟呼叫的系统和方法可以涉及检测功能的虚拟呼叫，并且使用该功能的单个调度来调用虚拟呼叫的两个或多个信道的虚拟呼叫。在一个示例中，确定两个或更多个信道共享公共目标地址，并且相对于公共目标地址进行该功能的单个调度。可以对共享共同目标地址的虚拟呼叫的附加信道重复该过程。

82. 发明申请

US20140253575A1 TRANSPOSE OF IMAGE DATA BETWEEN A LINEAR AND A Y-TILED STORAGE FORMAT 有权
标题翻译：线性和Y型倾斜存储格式之间的图像数据的传输
公开(公告)号：US20140253575A1
公开(公告)日：2014-09-11
申请号：US13977200
申请日：2011-12-29
申请人： Yuting Yang , Guei-Yuan Lueh , Stony Shen , John R. Hartwig , Kin-Hang Cheung
发明人： Yuting Yang , Guei-Yuan Lueh , Stony Shen , John R. Hartwig , Kin-Hang Cheung
IPC分类号： G06T1/60
CPC分类号： G06T1/60 , G06F3/14 , G09G5/363 , G09G5/397 , G09G2354/00 , G09G2360/121 , G09G2370/12 , H04N5/76
摘要： Systems, apparatus, articles, and methods are described including operations to transpose image data between a linear-type storage format and a Y-tiled-type storage format.
摘要翻译：描述了系统，装置，物品和方法，其包括在线性类型存储格式和Y平铺型存储格式之间转换图像数据的操作。

83. 发明授权

US08799876B2 Method and apparatus for assigning subroutines 有权
标题翻译：分配子程序的方法和装置
公开(公告)号：US08799876B2
公开(公告)日：2014-08-05
申请号：US11541083
申请日：2006-09-29
申请人： Guei-Yuan Lueh , Andrew T. Riffel , Hong Jiang , Bixia Zheng , Lian Tang
发明人： Guei-Yuan Lueh , Andrew T. Riffel , Hong Jiang , Bixia Zheng , Lian Tang
IPC分类号： G06F9/45
CPC分类号： G06F8/54
摘要： According to some embodiments, systems and methods are provided to link a first entry point of a first kernel to a dummy entry, link a second entry point of a second kernel to the dummy entry, and compile the first kernel and the second kernel.
摘要翻译：根据一些实施例，提供系统和方法来将第一内核的第一入口点连接到虚拟条目，将第二内核的第二入口点连接到虚拟条目，并且编译第一内核和第二内核。

84. 发明授权

US08633928B2 Reducing the bandwidth of sampler loads in shaders 有权
标题翻译：降低采样器负载在着色器中的带宽
公开(公告)号：US08633928B2
公开(公告)日：2014-01-21
申请号：US12540424
申请日：2009-08-13
申请人： Wei-Yu Chen , Guei-Yuan Lueh
发明人： Wei-Yu Chen , Guei-Yuan Lueh
IPC分类号： G06T15/50 , G06T15/60 , G06T1/00 , G06F15/00
CPC分类号： G06T15/005
摘要： Bandwidth may be conserved in messages transferred between a shader and a sampler by converting loads in array of structures format to structure of arrays format. More particularly, four four channel loads in array of structures format may be combined into a single eight channel message in structure of arrays format. Then, the eight channel message in structure of arrays format may be converted back to array of structures format.
摘要翻译：通过将结构数组中的负载转换为数组格式的结构，可以在着色器和采样器之间传输的消息中节省带宽。更具体地，结构格式阵列中的四个四通道负载可以组合成阵列格式结构的单个八通道消息。然后，数组格式结构中的八通道消息可以转换回结构格式的数组。

85. 发明申请

US20120210098A1 ENABLING VIRTUAL CALLS IN A SIMD ENVIRONMENT 有权
标题翻译：在SIMD环境中启用虚拟呼叫
公开(公告)号：US20120210098A1
公开(公告)日：2012-08-16
申请号：US13028574
申请日：2011-02-16
申请人： Wei-Yu Chen , Guei-Yuan Lueh , Subramaniam Maiyuran
发明人： Wei-Yu Chen , Guei-Yuan Lueh , Subramaniam Maiyuran
IPC分类号： G06F9/38
CPC分类号： G06F9/547 , G06F9/30061 , G06F9/449 , G06F9/455 , G06F15/8007
摘要： Systems and methods of enabling virtual calls in a single instruction multiple data (SIMD) environment may involve detecting a virtual call of a function and using a single dispatch of the function to invoke the virtual call for two or more channels of the virtual call. In one example, it is determined that the two or more channels share a common target address and a single dispatch of the function is conducted with respect to the common target address. The process may be iterated for additional channels of the virtual call that share a common target address.
摘要翻译：在单个指令多数据（SIMD）环境中启用虚拟呼叫的系统和方法可以涉及检测功能的虚拟呼叫，并且使用该功能的单个调度来调用虚拟呼叫的两个或多个信道的虚拟呼叫。在一个示例中，确定两个或更多个信道共享公共目标地址，并且相对于公共目标地址进行该功能的单个调度。可以对共享共同目标地址的虚拟呼叫的附加信道重复该过程。

86. 发明授权

US07350200B2 Method and system of controlling dynamically compiled native code size 有权
标题翻译：控制动态编译的本地代码大小的方法和系统
公开(公告)号：US07350200B2
公开(公告)日：2008-03-25
申请号：US09821640
申请日：2001-03-29
申请人： Guei-Yuan Lueh , Tatiana Shpeisman , Michal Cierniak , Cheng-Hsueh Andrew Hsieh
发明人： Guei-Yuan Lueh , Tatiana Shpeisman , Michal Cierniak , Cheng-Hsueh Andrew Hsieh
IPC分类号： G06F9/45
CPC分类号： G06F9/45516
摘要： Space occupied by native code associated with a first method and stored within a native code space is reclaimed by determining whether the native code space exceeds a threshold in response to the invocation of a second method. Once the determination is made, byte code is compiled into native code associated with the second method and the native code associated with the first method may be reclaimed. In one embodiment, this reclamation occurs in response to a determination that the threshold has been exceeded and that the first method is inactive. In another embodiment, reclamation occurs in response to a determination that the threshold has been exceeded and that the first method is cold.
摘要翻译：通过与第一方法相关联并存储在本地代码空间中的本地代码占据的空间通过响应于第二方法的调用来确定本机代码空间是否超过阈值来回收。一旦做出确定，字节码被编译成与第二方法相关联的本地代码，并且可以回收与第一方法相关联的本地代码。在一个实施例中，响应于已经超过阈值并且第一种方法不活动的确定而发生该回收。在另一个实施例中，响应于已经超过阈值并且第一种方法是冷的确定而发生回收。

87. 发明授权

US07120775B2 Inter-procedural allocation of stacked registers for a processor 有权
标题翻译：用于处理器的堆叠寄存器的程序间分配
公开(公告)号：US07120775B2
公开(公告)日：2006-10-10
申请号：US10747426
申请日：2003-12-29
申请人： Yang Liu , Sun Chan , Guangrong Gao , Dz-Ching (Roy) Ju , Guei-Yuan Lueh , Zhaoqing Zhang
发明人： Yang Liu , Sun Chan , Guangrong Gao , Dz-Ching (Roy) Ju , Guei-Yuan Lueh , Zhaoqing Zhang
IPC分类号： G06F12/00
CPC分类号： G06F8/441
摘要： A method for an allocation of stacked registers for Intel's Itanium® processor includes a three step process. Step I determines an intra-procedural stacked register usage by a program having a plurality of procedures. In step II, the disclosed method performs an inter-procedural analysis to assign quota of stacked register usage to every procedure. In step III, each procedure is allocated stacked register usage based on the quota assignments of step II.
摘要翻译：英特尔安腾（R）处理器分配堆叠寄存器的方法包括三步骤过程。步骤I通过具有多个过程的程序来确定程序内堆叠的寄存器使用。在步骤II中，所公开的方法执行程序间分析以将堆叠的寄存器使用的配额分配给每个过程。在步骤III中，基于步骤II的配额分配，分配每个过程的堆叠寄存器使用。

88. 发明授权

US06968546B2 Debugging support using dynamic re-compilation 失效
标题翻译：调试支持使用动态重新编译
公开(公告)号：US06968546B2
公开(公告)日：2005-11-22
申请号：US09822090
申请日：2001-03-30
申请人： Guei-Yuan Lueh
发明人： Guei-Yuan Lueh
IPC分类号： G06F9/45 , G06F11/36
CPC分类号： G06F11/3644 , G06F9/45516 , G06F11/3624
摘要： The present invention is a method and system to support debug. A function is re-compiled when a field watch for a field is activated. The function includes a byte code sequence having a field byte code that accesses or modifies the field. The re-compiled function provides a native code and occupies a code space. An instrumentation code corresponding to the field watch of the field is generated. The instrumentation code is inserted to the native code.
摘要翻译：本发明是一种支持调试的方法和系统。激活字段的字段监视时，会重新编译一个函数。该功能包括具有访问或修改该字段的字段字节码的字节码序列。重新编译的函数提供一个本地代码并占用一个代码空间。生成与场的视野对应的仪表代码。仪器代码被插入本机代码。

89. 发明申请

US20050144386A1 Mechanism to store reordered data with compression 失效
标题翻译：通过压缩来存储重新排序的数据的机制
公开(公告)号：US20050144386A1
公开(公告)日：2005-06-30
申请号：US10747470
申请日：2003-12-29
申请人： Ali-Reza Adl-Tabatabai , Anwar Ghuloum , Guei-Yuan Lueh , Victor Ying
发明人： Ali-Reza Adl-Tabatabai , Anwar Ghuloum , Guei-Yuan Lueh , Victor Ying
IPC分类号： G06F12/08 , H03M7/30 , G06F12/00
CPC分类号： G06F12/0802 , G06F12/0862 , G06F2212/401 , H03M7/30
摘要： According to one embodiment a computer system is disclosed. The computer system includes a central processing unit (CPU), a cache memory coupled to the CPU and a cache controller, coupled to the cache memory. The cache memory includes a plurality of compressible cache lines to store additional data. The cache controller reorders a cache line after each access to the cache line prior to the compression of the cache line into a compressed cache line.
摘要翻译：根据一个实施例，公开了一种计算机系统。计算机系统包括中央处理单元（CPU），耦合到CPU的高速缓冲存储器和耦合到高速缓冲存储器的高速缓存控制器。高速缓冲存储器包括多个可压缩的高速缓存线以存储附加数据。在将高速缓存行压缩到压缩高速缓存行之前，高速缓存控制器在对高速缓存行进行每次访问之后重新排序高速缓存行。

90. 发明授权

US06317869B1 Method of run-time tracking of object references in Java programs 有权
标题翻译： Java程序中对象引用的运行时跟踪方法
公开(公告)号：US06317869B1
公开(公告)日：2001-11-13
申请号：US09587233
申请日：2000-06-02
申请人： Ali-Reza Adl-Tabatabai , Guei-Yuan Lueh
发明人： Ali-Reza Adl-Tabatabai , Guei-Yuan Lueh
IPC分类号： G06F944
CPC分类号： G06F9/45516 , G06F12/0253 , Y10S707/99953 , Y10S707/99957
摘要： Many programming languages utilize reference pointers in computer code. Furthermore, some of these programming languages perform memory management in the form of garbage collection. Once such language is Java. During the execution of a garbage collection routine, the computer may need to locate all the variables containing reference values. The present invention introduces a method for run-time tracking of object references in computer code and determining which variables contain references to objects at garbage collection sites. The method of the present invention first creates a bit vector in memory. The bit vector is then initialized. Second, each variable declared in the computer program that may be used to store a reference value is assigned a unique bit within this bit vector. Each bit is maintained to indicate whether the variable it is assigned to is currently storing a reference value. Specifically, when a variable is assigned a reference value, the corresponding bit in the bit vector is set. When a variable is assigned a non-reference value, the corresponding bit in the bit vector is cleared.
摘要翻译：许多编程语言都使用计算机代码中的参考指针。此外，这些编程语言中的一些以垃圾收集的形式执行内存管理。一旦这样的语言是Java。在执行垃圾回收程序期间，计算机可能需要定位包含引用值的所有变量。本发明引入了一种用于计算机代码中对象引用的运行时跟踪的方法，并且确定哪些变量包含对垃圾收集站点上的对象的引用。本发明的方法首先在存储器中产生位向量。然后初始化位向量。第二，在计算机程序中声明的可用于存储参考值的每个变量在该位向量中被分配一个唯一的位。维持每个位以指示其被分配的变量当前是否存储参考值。具体地说，当变量被赋予参考值时，位向量中的相应位被置位。当变量分配了非参考值时，位向量中的相应位将被清除。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式