公开(公告)号：US06493715B1

公开(公告)日：2002-12-10

申请号：US09481177

申请日：2000-01-12

Applicant: Mark Robert Funk ,  Gera Goft ,  Dimitry Kloper ,  Shlomit Pinter ,  Esther Yeger-Lotem

Inventor： Mark Robert Funk ,  Gera Goft ,  Dimitry Kloper ,  Shlomit Pinter ,  Esther Yeger-Lotem

IPC: G06F1730

CPC classification number: G06F9/5061 ,  G06F2209/505 ,  Y10S707/922 ,  Y10S707/959 ,  Y10S707/966 ,  Y10S707/99944

Abstract: A method for controlling operation of a computer software application running on a plurality of computing entities, which are members of a group of mutually-linked computing entities running the application within a distributed computing system. The method includes receiving an indication of a change in membership of the group together with a reason for the change. A membership change message is delivered to the members, so as to inform the members of the change and of the reason for the change.

Abstract translation: 一种用于控制在多个计算实体上运行的计算机软件应用的操作的方法，所述计算机实体是在分布式计算系统内运行所述应用的一组相互联系的计算实体的成员。 该方法包括接收与该变化的原因一起的组的成员变化的指示。 会员变更信息传递给会员，以便向会员通知该变更以及更改的原因。

公开(公告)号：US06314561B1

公开(公告)日：2001-11-06

申请号：US08420540

申请日：1995-04-12

Applicant: Mark Robert Funk ,  Steven Raymond Kunkel ,  Mikko Herman Lipasti ,  Bilha Mendelson ,  Robert Ralph Roedinger ,  William Jon Schmidt

Inventor： Mark Robert Funk ,  Steven Raymond Kunkel ,  Mikko Herman Lipasti ,  Bilha Mendelson ,  Robert Ralph Roedinger ,  William Jon Schmidt

IPC: G06F945

CPC classification number: G06F8/4442 ,  G06F12/0862

Abstract: The data cache management mechanism of the present invention is created by an optimizing compiler. The optimizing compiler intelligently places non-blocking preload instructions into the instruction stream of the computer system so as to minimize both the frequency and detrimental effect of cache misses. The non-blocking preload instructions are placed into the instruction stream based on the existence of predictor constructs that foretell what information the processor will need and when it will need it. As a result, cache misses are either avoided entirely or reduced in severity.

Abstract translation: 本发明的数据高速缓存管理机制由优化编译器创建。 优化编译器智能地将非阻塞预加载指令放置到计算机系统的指令流中，以便最小化高速缓存未命中的频率和不利影响。 基于预测器处理器将需要哪些信息以及何时需要它的预测器构造的存在，将非阻塞预加载指令放入指令流中。 因此，高速缓存未命中或是完全避免或严重性降低。

公开(公告)号：US08312462B2

公开(公告)日：2012-11-13

申请号：US12551162

申请日：2009-08-31

Applicant: Lynn Keat Chung ,  Mark Robert Funk ,  Richard Karl Kirkman ,  Angela Mae Newton ,  Don Darrell Reed

Inventor： Lynn Keat Chung ,  Mark Robert Funk ,  Richard Karl Kirkman ,  Angela Mae Newton ,  Don Darrell Reed

IPC: G06F9/46 ,  G06F9/26

CPC classification number: G06F9/5033 ,  G06F9/4843 ,  G06F9/505 ,  G06F9/5066

Abstract: Disclosed is an apparatus, method, and program product for identifying and grouping threads that have interdependent data access needs. The preferred embodiment of the present invention utilizes two different constructs to accomplish this grouping. A Memory Affinity Group (MAG) is disclosed. The MAG construct enables multiple threads to be associated with the same node without any foreknowledge of which threads will be involved in the association, and without any control over the particular node with which they are associated. A Logical Node construct is also disclosed. The Logical Node construct enables multiple threads to be associated with the same specified node without any foreknowledge of which threads will be involved in the association. While logical nodes do not explicitly identify the underlying physical nodes comprising the system, they provide a means of associating particular threads with the same node and other threads with other node(s).

Abstract translation: 公开了一种用于识别和分组具有相互依赖的数据访问需求的线程的装置，方法和程序产品。 本发明的优选实施例利用两种不同的构造来完成该分组。 记录亲和组（MAG）被公开。 MAG构造使得多个线程能够与相同的节点相关联，而没有任何预先知道哪个线程将涉及关联，并且不对与它们相关联的特定节点进行任何控制。 还公开了逻辑节点构造。 逻辑节点构造使得多个线程可以与相同的指定节点相关联，而无需预先知道关联中涉及的线程。 虽然逻辑节点没有明确地识别包括系统的底层物理节点，但它们提供了将特定线程与其他节点相同的节点和其他线程相关联的手段。

公开(公告)号：US07266540B2

公开(公告)日：2007-09-04

申请号：US10793470

申请日：2004-03-04

Applicant: Lynn Keat Chung ,  Christopher Francois ,  Mark Robert Funk ,  Richard Karl Kirkman ,  Henry Joseph May ,  Don Darrell Reed ,  Kenneth Charles Vossen

Inventor： Lynn Keat Chung ,  Christopher Francois ,  Mark Robert Funk ,  Richard Karl Kirkman ,  Henry Joseph May ,  Don Darrell Reed ,  Kenneth Charles Vossen

IPC: G06F17/30

CPC classification number: G06F9/505 ,  G06F9/4843 ,  G06F9/5033 ,  G06F9/5066 ,  Y10S707/99931 ,  Y10S707/99945

Abstract: Disclosed is an apparatus, method, and program product for observing the nodal workload balance of the system on an ongoing basis, and for dynamically changing the preferred nodes of existing threads in order to improve nodal balance. Workload balance is ascertained on a nodal basis and then thread-based workload information is collected. If the detected imbalance persists, the thread-based information is used to change the assignment of preferred nodes to threads to improve nodal workload balance.

Abstract translation: 公开了一种用于在持续的基础上观察系统的节点工作负载的装置，方法和程序产品，并且用于动态地改变现有线程的优选节点以便改善节点平衡。 工作负载平衡在节点的基础上确定，然后收集基于线程的工作负载信息。 如果检测到的不平衡持续存在，则基于线程的信息用于将首选节点的分配更改为线程以提高节点工作负载平衡。

公开(公告)号：US20050198642A1

公开(公告)日：2005-09-08

申请号：US10793347

申请日：2004-03-04

Applicant: Lynn Keat Chung ,  Christopher Francois ,  Mark Robert Funk ,  Richard Karl Kirkman ,  Henry Joseph May ,  Don Darrell Reed ,  Kenneth Charles Vossen

Inventor： Lynn Keat Chung ,  Christopher Francois ,  Mark Robert Funk ,  Richard Karl Kirkman ,  Henry Joseph May ,  Don Darrell Reed ,  Kenneth Charles Vossen

IPC: G06F9/46 ,  G06F15/16 ,  G06F15/18

CPC classification number: G06F9/505

Abstract: Mechanism for Assigning Home Nodes to Newly Created Threads Disclosed is an apparatus, method, and program product for associating threads with the processing nodes of a multi-nodal computer system. The assignment is accomplished by considering both the relative work capacity of, and the relative amount of ongoing work assigned to, each node. Initiation Weights are used in the preferred embodiment to reflect the desired proportions of threads assigned to each node, from the standpoint of which node should be chosen when a new thread is created.

Abstract translation: 将主节点分配给新创建的线程的机制公开了一种用于将线程与多节点计算机系统的处理节点相关联的装置，方法和程序产品。 通过考虑分配给每个节点的正在进行的工作的相对工作能力和相对数量来完成任务。 在优选实施例中使用起始权重来反映分配给每个节点的线程的期望比例，从创建新线程时应选择哪个节点的角度来看。

公开(公告)号：US5784697A

公开(公告)日：1998-07-21

申请号：US622230

申请日：1996-03-27

Applicant: Mark Robert Funk ,  Larry Keith McMains ,  Donald Arthur Morrison ,  Robert Anthony Petrillo ,  Robert Carl Seemann ,  Arthur Douglas Smet ,  Timothy Joseph Torzewski

Inventor： Mark Robert Funk ,  Larry Keith McMains ,  Donald Arthur Morrison ,  Robert Anthony Petrillo ,  Robert Carl Seemann ,  Arthur Douglas Smet ,  Timothy Joseph Torzewski

IPC: G06F9/50 ,  G06F12/06 ,  G06F12/02 ,  G06F9/44

CPC classification number: G06F9/5016

Abstract: According to the present invention, pool allocation and process assignment mechanisms create process nodal affinity in a NUMA multiprocessor system for enhanced performance. The multiprocessor system includes multiple interconnected multiprocessing nodes that each contain one or more processors and a local main memory, the system main storage being distributed among the local main memories of the multiprocessing nodes in a NUMA architecture. A pool reservation mechanism reserves pools of memory space within the logical main storage, and the pool allocation mechanism allocates those pools to real pages in the local main-memory of multiprocessing nodes. Processes to be created on the multiprocessor are given an attribute that indicates an associated pool. Upon creation, the process assignment mechanism will only assign a process to a multiprocessing node that has been allocated the pool indicated by the process' attribute. This process nodal affinity increases accesses by the assigned process to local main storage of that node, thereby enhancing system performance.

Abstract translation: 根据本发明，池分配和过程分配机制在NUMA多处理器系统中创建过程节点亲和度以提高性能。 多处理器系统包括多个互连的多处理节点，每个节点包含一个或多个处理器和本地主存储器，该系统主存储器分布在NUMA架构中的多处理节点的本地主存储器之间。 池预留机制在逻辑主存储中保留存储空间池，池分配机制将这些池分配给多处理节点的本地主存中的实际页面。 在多处理器上创建的进程被赋予一个指示关联的池的属性。 创建后，进程分配机制将仅将一个进程分配给已分配由进程'属性指定的池的多处理节点。 此过程节点亲和性增加了分配过程对该节点的本地主存储的访问，从而提高系统性能。