International Business Machines Corporation

Don Darrell Reed

Richard Karl Kirkman

Lynn Keat Chung

Thomas M Gooding

Kenneth Charles Vossen

Alan G Gara

Henry Joseph May

Patrick Joseph McCarthy

Rudolf Adriaan Haring

Thomas A Liebsch

Paul William Coteus

Martin Ohmacht

Gerard Vincent Kopcsay

Mark Robert Funk

Christopher Francois

Dong Chen

Richard A Swetz

Angela Mae Newton

Chistopher Francois

Philip Heidelberger

Todd Edward Takken

Mark Edwin Giampapa

Burkhard Steinmacher-Burow

Mark Gregory Megerian

Mark James Jeanson

Todd Alan Inglett

Robert Matthew Senger

Noel A Eisley

Yutaka Sugawara

Paul Gregory Crumley

Matthew Roger Ellavsky

Thomas Edward Musta

Ralph E Bellofatto

Brandon Edward Schenck

Daniel J Littrell

Ralph A Bellofatto

Ross L Franke

An apparatus, method and computer program product for automatically controlling power dissipation of a parallel computing system that includes a plurality of processors. A computing device issues a command to the parallel computing system. A clock pulse-width modulator encodes the command in a system clock signal to be distributed to the plurality of processors. The plurality of processors in the parallel computing system receive the system clock signal including the encoded command, and adjusts power dissipation according to the encoded command.

System-wide power management control via clock distribution network

Fixing a problem is usually greatly aided if the problem is reproducible. To ensure reproducibility of a multiprocessor system, the following aspects are proposed: a deterministic system start state, a single system clock, phase alignment of clocks in the system, system-wide synchronization events, reproducible execution of system components, deterministic chip interfaces, zero-impact communication with the system, precise stop of the system and a scan of the system state.

Reproducibility in a multiprocessor system

A circuit generates a global clock signal with a pulse width modification to synchronize processors in a parallel computing system. The circuit may include a hardware module and a clock splitter. The hardware module may generate a clock signal and performs a pulse width modification on the clock signal. The pulse width modification changes a pulse width within a clock period in the clock signal. The clock splitter may distribute the pulse width modified clock signal to a plurality of processors in the parallel computing system.

Global synchronization of parallel processors using clock pulse width modulation

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).

Reducing remote memory accesses to shared data in a multi-nodal computer system

An apparatus, program product and method in which application program-specified resource allocation and allocation strength preferences are used to allocate hardware resources from a computer for that application program. The resource allocation preference is used to allocate hardware resources for the application program, while the allocation strength preference is used to determine whether to allocate alternate hardware resources when the resource allocation preference specified by the application program cannot be met.

Resource allocation in a NUMA architecture based on application specified resource and strength preferences for processor and memory resources

Resource allocation in a NUMA architecture based on separate application specified resource and strength preferences for processor and memory resources

An apparatus and method for controlling power usage in a computer includes a plurality of computers communicating with a local control device, and a power source supplying power to the local control device and the computer. A plurality of sensors communicate with the computer for ascertaining power usage of the computer, and a system control device communicates with the computer for controlling power usage of the computer.

Power throttling of collections of computing elements

Proactive power management in a parallel computer, the parallel computer including a service node and a plurality of compute nodes, the service node connected to the compute nodes through an out-of-band service network, each compute node including a computer processor and a computer memory operatively coupled to the computer processor. Embodiments include receiving, by the service node, a user instruction to initiate a job on an operational group of compute nodes in the parallel computer, the instruction including power management attributes for the compute nodes; setting, by the service node in accordance with the power management attributes for the compute nodes of the operational group, power consumption ratios for each compute node of the operational group including a computer processor power consumption ratio and a computer memory power consumption ratio; and initiating, by the service node, the job on the compute nodes of the operational group of the parallel computer.

Proactive power management in a parallel computer

Mechanism for reducing remote memory accesses to shared data in a multi-nodal computer system

Disclosed is an apparatus, method, and program product that enables distribution of operating system resources on a nodal basis in the same proportions as the expected system workload. The preferred embodiment of the present invention accomplishes this by assigning various types of weights to each node to represent their proportion of the overall balance within the system. Target Weights represent the desired distribution of the workload based on the existing proportions of processor and memory resources on each node. The actual workload balance on the system is represented by Current Weights, which the operating system strives to keep as close to the Target Weights as possible, on an ongoing basis. When the system is started, operating system services distribute their resources nodally in the same proportions as the Target Weights, and can request to be notified if the Target Weights ever change. If processors and/or memory are subsequently added or removed, new Target Weights are calculated at that time, and all services which requested notification are notified so they can redistribute their resources according to the new Target Weights or a stepwise refinement thereof.

Mechanism for enabling the distribution of operating system resources in a multi-node computer system

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.

Mechanism for dynamic workload rebalancing in a multi-nodal computer system

Growing community of inventors

Mantorville, MN, United States of America

Don Darrell Reed