-
公开(公告)号:US20110057937A1
公开(公告)日:2011-03-10
申请号:US12556401
申请日:2009-09-09
申请人: Ren Wu , Bin Zhang , Meichun Hsu
发明人: Ren Wu , Bin Zhang , Meichun Hsu
CPC分类号: G06F15/8023 , G06T1/00
摘要: A method is provided for optimizing computer processes executing on a graphics processing unit (GPU) and a central processing unit (CPU). Process data is subdivided into sequentially processed data and parallel processed data. The parallel processed data is subdivided into a plurality of data blocks assigned to a plurality of processing cores of the GPU. The data blocks on the GPU are processed with other data blocks in parallel on the plurality of processing cores. Sequentially processed data is processed on the CPU. Result data processed on the CPU is returned.
摘要翻译: 提供了一种用于优化在图形处理单元(GPU)和中央处理单元(CPU)上执行的计算机处理的方法。 过程数据被细分为顺序处理的数据和并行处理的数据。 并行处理的数据被细分为分配给GPU的多个处理核心的多个数据块。 GPU上的数据块在多个处理核上并行处理与其他数据块。 在CPU上处理顺序处理的数据。 返回CPU处理的结果数据。
-
12.
公开(公告)号:US07761444B2
公开(公告)日:2010-07-20
申请号:US11543494
申请日:2006-10-05
申请人: Bin Zhang , Qiming Chen , Umeshwar Dayal , Meichun Hsu
发明人: Bin Zhang , Qiming Chen , Umeshwar Dayal , Meichun Hsu
IPC分类号: G06F19/00
CPC分类号: G06F17/30442
摘要: A computer-executable method receives at least one query that specifies requests involving blocks of data. Based on information associated with at least one query, a sequence is identified relating to an order in which the blocks are to be retrieved into a storage for processing the requests. An output is provided for the at least one query based on retrieving the blocks into the storage according to the identified sequence. According to another implementation, a method includes caching blocks of data associated with processing a database query in a temporary storage, and replacing at least some of the blocks of data in the temporary storage using a first replacement algorithm this is based on information associated with processing the database query, and where the first algorithm is different from a second replacement algorithm used by an operating system of the database system.
摘要翻译: 计算机可执行方法接收至少一个指定涉及数据块的请求的查询。 基于与至少一个查询相关联的信息,识别与要将数据块检索到用于处理请求的存储器中的顺序有关的序列。 基于根据所识别的序列将块检索到存储器中,为至少一个查询提供输出。 根据另一实施方式,一种方法包括缓存与临时存储器中处理数据库查询相关联的数据块,并且使用基于与处理相关联的信息的第一替换算法来替换临时存储器中的至少一些数据块 数据库查询,以及第一种算法与数据库系统的操作系统使用的第二种替换算法的不同之处。
-
公开(公告)号:US06944607B1
公开(公告)日:2005-09-13
申请号:US09684488
申请日:2000-10-04
申请人: Bin Zhang , Meichun Hsu , Igor Kleyner
发明人: Bin Zhang , Meichun Hsu , Igor Kleyner
IPC分类号: G06F7/00
CPC分类号: G06K9/6223 , Y10S707/915 , Y10S707/968 , Y10S707/99932 , Y10S707/99945
摘要: An aggregated data clustering method and system. First, the data points to be clustered and a size parameter are received. The size parameter specifies the number of data points to be moved at one time in the clustering algorithm. Next, the data points are clustered by using an aggregated clustering algorithm (e.g., aggregated local K-Means clustering algorithm) and the size parameter to generate clustered results. Then, a determination is made whether of not the clustered results are satisfactory. If the clustered results are satisfactory, the clustering is stopped. Otherwise, a modified or refined parameter size is received. Then, clustering is performed on the clustered results generated previously by using the aggregated clustering algorithm and the revised or refined parameter size. The steps of determining, modifying the parameter size, and aggregated clustering are repeated until satisfactory clustering results are achieved.
摘要翻译: 聚合数据聚类方法和系统。 首先,要收集要聚类的数据点和大小参数。 size参数指定在聚类算法中一次移动的数据点的数量。 接下来,通过使用聚合聚类算法(例如聚合的本地K均值聚类算法)和大小参数来聚集数据点以生成聚类结果。 然后,确定聚类结果是否满意。 如果聚类结果令人满意,则停止聚类。 否则,接收修改或改进的参数大小。 然后,通过使用聚合聚类算法和修改或精化的参数大小对先前生成的聚类结果执行聚类。 重复确定,修改参数大小和聚合聚类的步骤,直到达到令人满意的聚类结果。
-
公开(公告)号:US08799192B2
公开(公告)日:2014-08-05
申请号:US13406843
申请日:2012-02-28
申请人: Bin Zhang , Meichun Hsu
发明人: Bin Zhang , Meichun Hsu
IPC分类号: G06F15/18
CPC分类号: G06F17/30958
摘要: A nested chain of densest subgraphs is derived by a computer from a given graph that has multiple vertices and edges. The two ends of each edge are assigned with respective incident weights, and each vertex is given a vertex weight. A weight balancing process is carried out by the computer to iteratively go through the edges to adjust the incident weights of each edge and the vertex weights of the vertices connected by that edge to reduce a difference between the vertex weights of the two vertices. After the balancing, the vertex weights are put in an ordered sequence according to their values, and a nested chain of densest subgraphs is derived from the ordered sequence.
摘要翻译: 密集子图的嵌套链由计算机从具有多个顶点和边的给定图导出。 每个边缘的两端被分配有相应的事件权重,并且每个顶点被赋予顶点权重。 计算机进行权重平衡处理,以迭代地遍历边缘,以调整每个边缘的入射权重和由该边连接的顶点的顶点权重,以减少两个顶点的顶点权重之间的差异。 在平衡之后,顶点权重根据它们的值被置于有序序列中,并且从有序序列导出嵌套的密集子图链。
-
公开(公告)号:US08751205B2
公开(公告)日:2014-06-10
申请号:US13172553
申请日:2011-06-29
申请人: Meichun Hsu , Qiming Chen , Bin Zhang
发明人: Meichun Hsu , Qiming Chen , Bin Zhang
IPC分类号: G06G7/48
CPC分类号: G06F17/30516 , G06F17/30289
摘要: A method of generating discrete event simulation data using a query is provided herein. The method includes retrieving a trajectory specifying a service access event for a customer accessing a service in a simulation. The method also includes processing, in a relational database query, one or more pending events corresponding to one or more customers accessing the service before the customer. Additionally, the method includes processing, in the relational database query, the service access event. Processing the service access event occurs after processing the one or more pending events. The method also includes providing a result for the relational database query. The result comprises the trajectory, and the one or more trajectories.
摘要翻译: 本文提供了使用查询产生离散事件仿真数据的方法。 该方法包括检索在模拟中为访问服务的客户指定服务访问事件的轨迹。 该方法还包括在关系数据库查询中处理与客户之前访问服务的一个或多个客户相对应的一个或多个未决事件。 另外,该方法包括在关系数据库查询中处理服务访问事件。 在处理一个或多个未决事件之后,处理服务访问事件。 该方法还包括提供关系数据库查询的结果。 结果包括轨迹和一个或多个轨迹。
-
公开(公告)号:US20130080474A1
公开(公告)日:2013-03-28
申请号:US13246771
申请日:2011-09-27
申请人: Bin Zhang , Meichun Hsu
发明人: Bin Zhang , Meichun Hsu
IPC分类号: G06F7/00
CPC分类号: G06F16/24566
摘要: A computer apparatus and related method to accelerate recursive queries is provided. In one aspect, a node located within a data structure may be accessed. The data structure may emulate a hierarchy of interlinked nodes. In another aspect, a series of bits may be generated that indicate a level of the node within the hierarchy and that outline a unique path of interlinked nodes toward the node.
摘要翻译: 提供了一种用于加速递归查询的计算机装置和相关方法。 在一个方面,可以访问位于数据结构内的节点。 数据结构可以模拟互连节点的层次结构。 在另一方面,可以生成一系列比特,其指示层次结构内的节点的级别,并且向节点轮廓形成相互连接的节点的唯一路径。
-
公开(公告)号:US08400458B2
公开(公告)日:2013-03-19
申请号:US12556401
申请日:2009-09-09
申请人: Ren Wu , Bin Zhang , Meichun Hsu
发明人: Ren Wu , Bin Zhang , Meichun Hsu
IPC分类号: G06F15/80
CPC分类号: G06F15/8023 , G06T1/00
摘要: A method is provided for optimizing computer processes executing on a graphics processing unit (GPU) and a central processing unit (CPU). Process data is subdivided into sequentially processed data and parallel processed data. The parallel processed data is subdivided into a plurality of data blocks assigned to a plurality of processing cores of the GPU. The data blocks on the GPU are processed with other data blocks in parallel on the plurality of processing cores. Sequentially processed data is processed on the CPU. Result data processed on the CPU is returned.
摘要翻译: 提供了一种用于优化在图形处理单元(GPU)和中央处理单元(CPU)上执行的计算机处理的方法。 过程数据被细分为顺序处理的数据和并行处理的数据。 并行处理的数据被细分为分配给GPU的多个处理核心的多个数据块。 GPU上的数据块在多个处理核上并行处理与其他数据块。 在CPU上处理顺序处理的数据。 返回CPU处理的结果数据。
-
18.
公开(公告)号:US20080086446A1
公开(公告)日:2008-04-10
申请号:US11543494
申请日:2006-10-05
申请人: Bin Zhang , Qiming Chen , Umeshwar Dayal , Meichun Hsu
发明人: Bin Zhang , Qiming Chen , Umeshwar Dayal , Meichun Hsu
IPC分类号: G06F17/30
CPC分类号: G06F17/30442
摘要: A computer-executable method receives at least one query that specifies requests involving blocks of data. Based on information associated with at least one query, a sequence is identified relating to an order in which the blocks are to be retrieved into a storage for processing the requests. An output is provided for the at least one query based on retrieving the blocks into the storage according to the identified sequence.
摘要翻译: 计算机可执行方法接收至少一个指定涉及数据块的请求的查询。 基于与至少一个查询相关联的信息,识别与要将数据块检索到用于处理请求的存储器中的顺序有关的序列。 基于根据所识别的序列将块检索到存储器中,为至少一个查询提供输出。
-
公开(公告)号:US07039638B2
公开(公告)日:2006-05-02
申请号:US09844730
申请日:2001-04-27
申请人: Bin Zhang , Meichun Hsu , George Forman
发明人: Bin Zhang , Meichun Hsu , George Forman
IPC分类号: G06F17/30
CPC分类号: G06F17/30705 , G06K9/00993 , G06K9/6218 , Y10S707/99937 , Y10S707/99943
摘要: A distributed data clustering system having an integrator and at least two computing units. Each computing unit is loaded with common global parameter values and a particular local data set. Each computing unit then generates local sufficient statistics based on the local data set and global parameter values. The integrator employs the local sufficient statistics of all the computing units to update the global parameter values.
摘要翻译: 一种具有积分器和至少两个计算单元的分布式数据聚类系统。 每个计算单元加载有公共全局参数值和特定的本地数据集。 然后,每个计算单元基于本地数据集和全局参数值生成本地足够的统计信息。 集成商使用所有计算单元的本地足够的统计信息来更新全局参数值。
-
公开(公告)号:US09760372B2
公开(公告)日:2017-09-12
申请号:US13224090
申请日:2011-09-01
申请人: Ren Wu , Bin Zhang , Meichun Hsu , Qiming Chen
发明人: Ren Wu , Bin Zhang , Meichun Hsu , Qiming Chen
CPC分类号: G06F9/30036 , G06F9/3001 , G06F9/5066
摘要: A method for combining data values through associative operations. The method includes, with a processor, arranging any number of data values into a plurality of columns according to natural parallelism of the associative operations and reading each column to a register of an individual processor. The processors are directed to combine the data values in the columns in parallel using a first associative operation. The results of the first associative operation for each column are stored in a register of each processor.
-
-
-
-
-
-
-
-
-