Author pages are created from data sourced from our academic publisher partnerships and public sources.
Share This Author
TimeGraph: GPU Scheduling for Real-Time Multi-Tasking Environments
- S. Kato, Karthik Lakshmanan, R. Rajkumar, Y. Ishikawa
- Computer ScienceUSENIX Annual Technical Conference
- 15 June 2011
TimeGraph is presented, a real-time GPU scheduler at the device-driver level for protecting important GPU workloads from performance interference and supports two priority-based scheduling policies in order to address the tradeoff between response times and throughput introduced by the asynchronous and non-preemptive nature of GPU processing.
The International Exascale Software Project roadmap
- J. Dongarra, P. Beckman, K. Yelick
- Computer ScienceInt. J. High Perform. Comput. Appl.
- 1 February 2011
The work of the community to prepare for the challenges of exascale computing is described, ultimately combing their efforts in a coordinated International Exascale Software Project.
Gang EDF Scheduling of Parallel Task Systems
An algorithm, called Gang EDF, which applies the Earliest Deadline First (EDF) policy to the traditional Gang scheduling scheme is presented and new ideas for the parallel task model are introduced.
Semi-partitioned Scheduling of Sporadic Task Systems on Multiprocessors
- S. Kato, N. Yamasaki, Y. Ishikawa
- Computer Science21st Euromicro Conference on Real-Time Systems
- 1 July 2009
A new algorithm for scheduling of sporadic task systems with arbitrary deadlines on identical multiprocessor platforms based on the concept of semi-partitioned scheduling, in which most tasks are fixed to specific processors, while a few tasks migrate across processors.
RGEM: A Responsive GPGPU Execution Model for Runtime Engines
- S. Kato, Karthik Lakshmanan, Aman Kumar, Mihir Kelkar, Y. Ishikawa, R. Rajkumar
- Computer ScienceIEEE 32nd Real-Time Systems Symposium
- 29 November 2011
The experiments demonstrate that the response times of high-priority GPGPU tasks can be protected under RGEM, whereas their response times increase in an unbounded fashion without RGEM support, as the data sizes of competing workload increase.
Al-1/d: a distributed programming system with multi-model reflection framework
The application of Multi-Model Re ection Framework to a distributed system is reported and the requirement of the network transparency, by which the user do not need to take the nature of the distributed environment into account, is reported.
Scalable Distributed Monte-Carlo Tree Search
- Kazuki Yoshizoe, Akihiro Kishimoto, Tomoyuki Kaneko, H. Yoshimoto, Y. Ishikawa
- Computer ScienceSOCS
- 5 July 2011
This work presents a massively parallel MCTS algorithm, that applies the TDS parallelism to the Upper Confidence bound Applied to Trees (UCT) algorithm, which is the most representative M CTS algorithm and introduces a reformulation of UCT called Depth-First UCT to drastically decrease communication overhead.
Resource Sharing in GPU-Accelerated Windowing Systems
- S. Kato, Karthik Lakshmanan, Y. Ishikawa, R. Rajkumar
- Computer Science17th IEEE Real-Time and Embedded Technology and…
- 11 April 2011
This paper proposes two protocols that enable application tasks to efficiently share the GPU resource in the X Window System and identifies and addresses resource-sharing problems raised in GPU-accelerated windowing systems.
Design and Evaluation of Precise Software Pacing Mechanisms for Fast Long-Distance Networks
An IPGaware packet scheduling mechanism is proposed, in order to merge multiple streams which go through each of different bottleneck links, where the pacing ratios are appropriately maintained.
EZTrace: A Generic Framework for Performance Analysis
- François Trahay, François Rué, Mathieu Faverge, Y. Ishikawa, R. Namyst, J. Dongarra
- Computer Science11th IEEE/ACM International Symposium on Cluster…
- 23 May 2011
Eztrace, a generic trace generation framework that aims at providing a simple way to analyze applications, is presented, based on plugins that allow it to trace different programming models such as MPI, pthread or OpenMP as well as user-defined libraries or applications.