International Business Machines Corporation

Nec Corporation

Nec Laboratories America, Inc.

Cheng-Hong Li

Srimat T Chakradhar

Giuseppe Coviello

Srihari Cadambi

Rajat Phull

Kunal Rao

Eugen Schenfeld

Alfred J Park

Murugan Sankaradas

Murugan Sankaradass

Arash Rezaei

Konstantinos Katrinis

Shoukat Ali

Deepak Ajwani

Yoonseo Choi

An embodiment of the invention includes applying a first partition to a plurality of LPs, wherein a particular LP is assigned to a first set of LPs. A second partition is applied to the LPs, wherein the particular LP is assigned to an LP set different from the first set. For both the first and second partitions, lookahead values and transit times are determined for each of the LPs and related links. For the first partition, a first system progression rate is computed using a specified function with the lookahead values and transit times determined for the first partition. For the second partition, a second system progression rate is computed using the specified function with the lookahead values and transit times determined for the second partition. The first and second system progression rates are compared to determine which is the lowest.

Workload partitioning procedure for null message-based PDES

Methods are provided. A method includes capturing a snapshot of an offload process being executed by one or more many-core processors. The offload process is in signal communication with a host process being executed by a host processor. At least the offload is in signal communication with a monitoring process. The method further includes terminating the offload process on the one or more many-core processors, by the monitor process responsive to a communication between the monitor process and the offload processing being disrupted. The snapshot includes a respective predetermined minimum set of information required to restore a same state of the process as when the snapshot was taken.

Capturing snapshots of offload applications on many-core coprocessors

Systems and methods for swapping out and in pinned memory regions between main memory and a separate storage location in a system, including establishing an offload buffer in an interposing library; swapping out pinned memory regions by transferring offload buffer data from a coprocessor memory to a host processor memory, unregistering and unmapping a memory region employed by the offload buffer from the interposing library, wherein the interposing library is pre-loaded on the coprocessor, and collects and stores information employed during the swapping out. The pinned memory regions are swapped in by mapping and re-registering the files to the memory region employed by the offload buffer, and transferring data of the offload buffer data from the host memory back to the re-registered memory region.

Systems and methods for swapping pinned memory buffers

Methods are provided. A method for swapping-out an offload process from a coprocessor includes issuing a snapify_pause request from a host processor to the coprocessor to initiate a pausing of the offload process executing by the coprocessor and another process executing by the host processor using a plurality of locks. The offload process is previously offloaded from the host processor to the coprocessor. The method further includes issuing a snapify_capture request from the host processor to the coprocessor to initiate a local snapshot capture and saving of the local snapshot capture by the coprocessor. The method also includes issuing a snapify_wait request from the host processor to the coprocessor to wait for the local snapshot capture and the saving of the local snapshot capture to complete by the coprocessor.

Methods and systems for scheduling jobs to manycore nodes in a cluster include selecting a job to run according to the job's wait time and the job's expected execution time; sending job requirements to all nodes in a cluster, where each node includes a manycore processor; determining at each node whether said node has sufficient resources to ever satisfy the job requirements and, if no node has sufficient resources, deleting the job; creating a list of nodes that have sufficient free resources at a present time to satisfy the job requirements; and assigning the job to a node, based on a difference between an expected execution time and associated confidence value for each node and a hypothetical fastest execution time and associated hypothetical maximum confidence value.

Simultaneous scheduling of processes and offloading computation on many-core coprocessors

A method is disclosed to manage a multi-processor system with one or more manycore devices, by managing real-time bag-of-tasks applications for a cluster, wherein each task runs on a single server node, and uses the offload programming model, and wherein each task has a deadline and three specific resource requirements: total processing time, a certain number of manycore devices and peak memory on each device; when a new task arrives, querying each node scheduler to determine which node can best accept the task and each node scheduler responds with an estimated completion time and a confidence level, wherein the node schedulers use an urgency-based heuristic to schedule each task and its offloads; responding to an accept/reject query phase, wherein the cluster scheduler send the task requirements to each node and queries if the node can accept the task with an estimated completion time and confidence level; and scheduling tasks and offloads using a aging and urgency-based heuristic, wherein the aging guarantees fairness, and the urgency prioritizes tasks and offloads so that maximal deadlines are met.

Method for simultaneous scheduling of processes and offloading computation on many-core coprocessors

Systems and methods are disclosed that share coprocessor resources between two or more applications in a computing cluster using a job selector to receive jobs from a job queue; a node selector coupled to the job selector; an off line profiler with an interference prediction model; a coprocessor dynamic interference detection module; and a coprocessor interference response module.

Interference-driven resource management for GPU-based heterogeneous clusters

A runtime method is disclosed that dynamically sets up core containers and thread-to-core affinity for processes running on manycore coprocessors. The method is completely transparent to user applications and incurs low runtime overhead. The method is implemented within a user-space middleware that also performs scheduling and resource management for both offload and native applications using the manycore coprocessors.

Methods of processing core selection for applications on manycore processors

A method, an apparatus and an article of manufacture for generating a synthetic workload for stream computing. The method includes generating an undirected graph that meets a node degree distribution parameter, obtaining a user-defined parameter for at least one feature for at least one stream computing application, and manipulating the undirected graph to generate a synthetic workload for stream computing in compliance with the user-defined parameter for the at least one feature for the at least one stream computing application.

Simulating stream computing systems

A method of detecting a bottleneck in a data flow program includes executing the data flow program, wherein the data flow program comprises a kernel, determining a percentage of time the kernel spends on a computation, and determining the kernel to be a bottleneck upon comparing the percentage of time the kernel spends on the computation to a condition.

Growing community of inventors

Princeton, NJ, United States of America

Cheng-Hong Li