Abstract: A process migration method comprising executing a computer program using a group of parallel processes, each process carrying out a computation, the execution using current computing resources to provide current group data as a result of the computations, deciding to change the resources, and making a choice between increasing the resources; decreasing the resources; and moving to different resources, wherein moving to different resources can include increase, decrease or maintenance of the resources. The method comprising communication between the current computing resources and changed computing resources to allow the program to execute on the changed resources, the communication comprising migration of the execution to changed resources and synchronization of migrated group data with the current group data; wherein execution using the current resources overlaps in time with the communication.

Abstract: A controller to instruct execution in an environment of plural computing resources. The controller comprising: an information collecting unit to collect available resource information of computing resources available to execute an application indicating an amount and/or type of computing resource available in categories of computing resource; scalability information including an indication of application execution rate; and performance target information including an indication of performance targets. The controller further comprises: a configuration selection unit to select a configuration which will come closest to meeting, the performance targets; and an instructing unit to instruct the execution of the application using the selected configuration.

Abstract: A synchronization method in a computer system with multiple cores, wherein a group of threads executes in parallel on a plurality of cores, the group of threads being synchronized using barrier synchronization in which each thread in the group waits for all the others at a barrier before progressing; the group of threads executes until a first thread reaches the barrier; the first thread enters a polling state, repeatedly checking for a release condition indicating the end of the barrier; subsequent threads to reach the barrier are moved to the core on which the first thread is executing; and other cores are powered down as the number of moved threads increases; and wherein when the first thread detects the release condition, the powered down cores are powered up and are available for use by the threads.

Abstract: A method of recovering application data from the memory of a failed node in a computer system comprising a plurality of nodes connected by an interconnect and of writing the application data to a replacement node; wherein a node of the computer system executes an application which creates application data storing the most recent state of the application in a node memory; the node fails; the node memory of the failed node is then controlled using a failover memory controller; and the failover memory controller copies the application data from the node memory of the failed node to a node memory of the replacement node over the interconnect.

Abstract: A process migration method comprising executing a computer program on current computing resources, in a current partition of computations to parallel processes, each parallel process carrying out at least one computation; communicating current partition data to intermediate computing resources; using the intermediate computing resources to calculate new partition data including a new allocation of computations to processes for use on changed computing resources; and migrating the process execution to the changed computing resources by communicating the new partition data to the changed computing resources.

Abstract: A method of injecting hardware faults into execution of an application in a distributed computing system comprising hardware components including linked nodes, the method comprising: loading an enhanced software stack allowing faults to be injected by deactivating or degrading hardware components as a result of fault triggers; running a fault-trigger daemon on each of the nodes; providing the fault trigger for a degradation or deactivation using one of the daemons to trigger a layer of the software stack controlling a hardware component to inject a fault into the hardware component; and continuing execution of the application with the injected fault.

Abstract: A method of improving fault tolerance in a computing system arranged to find a computational solution, the method comprising: computing at least two versions of the solution by using a hierarchy of at least two different solvers in parallel; and if there is a fault during execution of a solver resulting in a missing value, substituting a value from a solver that is lower in the hierarchy to replace the missing value.

Abstract: A process migration method comprising executing a computer program using a group of parallel processes, each process carrying out a computation, the execution using current computing resources to provide current group data as a result of the computations, deciding to change the resources, and making a choice between increasing the resources; decreasing the resources; and moving to different resources, wherein moving to different resources can include increase, decrease or maintenance of the resources. The method comprising communication between the current computing resources and changed computing resources to allow the program to execute on the changed resources, the communication comprising migration of the execution to changed resources and synchronization of migrated group data with the current group data; wherein execution using the current resources overlaps in time with the communication.

Abstract: A controller to instruct execution in an environment of plural computing resources. The controller comprising: an information collecting unit to collect available resource information of computing resources available to execute an application indicating an amount and/or type of computing resource available in categories of computing resource; scalability information including an indication of application execution rate; and performance target information including an indication of performance targets. The controller further comprises: a configuration selection unit to select a configuration which will come closest to meeting, the performance targets; and an instructing unit to instruct the execution of the application using the selected configuration.

Abstract: A synchronization method in a computer system with multiple cores, wherein a group of threads executes in parallel on a plurality of cores, the group of threads being synchronised using barrier synchronisation in which each thread in the group waits for all the others at a barrier before progressing; the group of threads executes until a first thread reaches the barrier; the first thread enters a polling state, repeatedly checking for a release condition indicating the end of the barrier; subsequent threads to reach the barrier are moved to the core on which the first thread is executing; and other cores are powered down as the number of moved threads increases; and wherein when the first thread detects the release condition, the powered down cores are powered up and are available for use by the threads.

Abstract: A method of recovering application data from the memory of a failed node in a computer system comprising a plurality of nodes connected by an interconnect and of writing the application data to a replacement node; wherein a node of the computer system executes an application which creates application data storing the most recent state of the application in a node memory; the node fails; the node memory of the failed node is then controlled using a failover memory controller; and the failover memory controller copies the application data from the node memory of the failed node to a node memory of the replacement node over the interconnect.

Abstract: An ultrasound probe device including an array of transducers, at least one of which is movable relative to at least one of the other transducers. Each movable transducer may be movable relative to at least one of the other transducers along one axis of movement only, preferably parallel with an initial direction of travel of emitted ultrasound waves relative to the device.

Abstract: A method of injecting hardware faults into execution of an application in a distributed computing system comprising hardware components including linked nodes, the method comprising: loading an enhanced software stack allowing faults to be injected by deactivating or degrading hardware components as a result of fault triggers; running a fault-trigger daemon on each of the nodes; providing the fault trigger for a degradation or deactivation using one of the daemons to trigger a layer of the software stack controlling a hardware component to inject a fault into the hardware component; and continuing execution of the application with the injected fault.

Abstract: A process migration method comprising executing a computer program on current computing resources, in a current partition of computations to parallel processes, each parallel process carrying out at least one computation; communicating current partition data to intermediate computing resources; using the intermediate computing resources to calculate new partition data including a new allocation of computations to processes for use on changed computing resources; and migrating the process execution to the changed computing resources by communicating the new partition data to the changed computing resources.

Abstract: A method of improving fault tolerance in a computing system arranged to find a computational solution, the method comprising: computing at least two versions of the solution by using a hierarchy of at least two different solvers in parallel; and if there is a fault during execution of a solver resulting in a missing value, substituting a value from a solver that is lower in the hierarchy to replace the missing value.

Abstract: The present invention relates to an ultrasound probe device. A device 100 embodying the present invention comprises an array of transducers 102, at least one of which is movable relative to at least one of the other transducers 102. Preferably, the or each movable transducer may be movable relative to at least one of the other transducers along one axis of movement only. This axis of movement is preferably parallel with an initial direction of travel of emitted ultrasound waves relative to the device.