Showing posts with label runtime systems. Show all posts
Showing posts with label runtime systems. Show all posts

Tuesday, April 12, 2016

Research Readings in Cloud

Using the Xeon Phi Platform to Run Speculatively-Parallelized Codes

A Estebanez, DR Llanos, A Gonzalez-Escribano - International Journal of Parallel …, 2016
Page 1. Int J Parallel Prog DOI 10.1007/s10766-016-0421-x Using the Xeon Phi
Platform to Run Speculatively-Parallelized Codes Alvaro Estebanez1 · Diego R.
Llanos1 · Arturo Gonzalez-Escribano1 Received: 14 July 2015 ...

[PDF] A Hierarchical Load Balancing Strategy Considering Communication Delay Overhead for Large Distributed Computing Systems

J Yang, L Ling, H Liu
Page 1. MATHEMATICAL PROBLEMS IN ENGINEERING, VOL. #, NO. #, MMMMMMMM 2016
1 A Hierarchical Load Balancing Strategy Considering Communication Delay Overhead for Large
Distributed Computing Systems Jixiang Yang 1 , Ling Ling 2 , and Haibin Liu 3 ...

Tuesday, April 5, 2016

Research Readings in Cloud

SYSTEMS MANAGEMENT BASED ON SEMANTIC MODELS AND LOW-LEVEL RUNTIME STATE

T Eilam, FA Oliveira, F Rosenberg - US Patent 20,160,077,816, 2016
... As will be discussed in greater detail below, the systems manager 110 provides improved and
enhanced systems management by correlating high-level system models, such as the semantic
models 112, and low-level runtime systems states as captured by the introspection ...

META: middleware for events, transactions, and analytics

M Arnold, D Grove, B Herta, M Hind, M Hirzel… - IBM Journal of Research …, 2016
Page 1. META: Middleware for Events, Transactions, and Analytics M. Arnold D. Grove
B. Herta M. Hind M. Hirzel A. Iyengar L. Mandel VA Saraswat A. Shinnar J. Siméon
M. Takeuchi O. Tardieu W. Zhang Businesses that receive ...

Notification of Blocking Tasks

T Kumar, P Montesinos Ortego, A Raman - US Patent 20,160,078,246, 2016
... Modern parallel programming runtime systems allow software designers to create
high-performance parallel application programs that better exploit the concurrency
and/or parallelism capabilities of modern processor architectures. ...

[PDF] A Survey on Hardware and Software Support for Thread Level Parallelism

S Mazumdar, R Giorgi - arXiv preprint arXiv:1603.09274, 2016
... S. Mazumdar et al. [EGS06], Chapel [CCZ07] are well-known PGAS programming languages.
PGAS also has very efficient runtime systems such as Global-Address Space Networking (GAS-
Net) [BJ02]. GASNet is used to exchange the active massages among the threads. ...

Monday, April 4, 2016

Research Readings in Cloud

Coordinated Garbage Collection in Distributed Systems

TL Harris, MC Maas - US Patent 20,160,070,593, 2016
... As described herein, it has been demonstrated that systems software and language runtime
systems may be able to evolve in a manner that better supports “rack scale” machines in which
tightly-coupled sets of machines are deployed and used together as a single system. ...

[PDF] Discovering and exploiting parallelism in DOACROSS loops

N Murphy - 2016
Page 1. Technical Report Number 882 Computer Laboratory UCAM-CL-TR-882 ISSN 1476-2986
Discovering and exploiting parallelism in DOACROSS loops Niall Murphy March 2016 15 JJ
Thomson Avenue Cambridge CB3 0FD United Kingdom phone +44 1223 763500 ...

[PDF] Ensemble Toolkit: Scalable and Flexible Execution of Ensembles of Tasks

V Balasubramanian, A Treikalis, O Weidner, S Jha
... 1) Job submission language: The Ensemble toolkit is de- signed to be compatible withruntime
systems that comply with the SAGA Job Description [39] that is derived from the standard Job
Submission Description Language [40] proposed by the Global Grid Forum. ...

[PDF] Modules, Componentization, and Transition

GDRP Curtis
... once produced. Similarly, an IFC can be edited without requiring a C++ compiler.
This is obviously useful for “smart” build systems and other development tools,
including IDEs, debuggers, runtime systems, etc. Note that an ..

Friday, April 1, 2016

Research Readings in Cloud

[PDF] Reducing the Complexity of Heterogeneous Computing: A Unified Approach for Application Development and Runtime Optimization

M Kicherer
... Therefore, runtime systems are used to dynamically determine and select the best implementation
and processing unit for a task in the current system. Similar to the runtime system in this work,
most of them use an empirical approach to determine the best combination. ...

Tuesday, March 29, 2016

Research Readings in Cloud

[PDF] Variation Among Processors Under Turbo Boost in HPC Systems

B Acun, P Miller, LV Kale
Page 1. Variation Among Processors Under Turbo Boost in HPC Systems Bilge Acun, Phil Miller,
Laxmikant V. Kale Department of Computer Science University of Illinois at Urbana-Champaign,
Urbana, IL, 61801 {acun2, mille121, kale}@illinois.edu Abstract ...

[PDF] Systemwide Power Management with Argo

D Ellsworth, T Patki, S Perarnau, S Seo, A Amer…
... Additionally, runtime systems and other mecha- nisms for optimizing application performance
under a power bound are being proposed [5], [6], [13], [14], [17], [27], [28]. ... Argobots as well as
other power-aware runtime systems provide such tuning opportunities at the ERM level. ...

[PDF] RDBG: a Reactive Programs Extensible Debugger

E Jahier
... In this article, we present RDBG, a debugger for reactive programs that re-use the LURETTE
plumbing and its ability to plug various kinds of reactive programs together in closed loop. RDBG
has a plugin mechanism to connect to reactive program runtime systems. ...

Monday, March 28, 2016

Research Readings in Cloud

[PDF] Timing Analysis of Automotive Architectures and Software

N Navet - 2016
Page 1. Timing Analysis of Automotive Architectures and Software Nicolas Navet University of
Luxembourg, founder RealTime-at-Work Page 2. Outline Nicolas Navet / University of Luxembourg
1 16-Mar-16 ✓Focus is not the formalisms but on what to expect from timing analysis ...

Friday, March 25, 2016

Research Readings in Cloud

A framework for accelerating local feature extraction with OpenCL on multi-core CPUs and co-processors

K Moren, D Göhringer - Journal of Real-Time Image Processing, 2016
... compiler systems. 4.1 OpenCL compiler and runtime systems OpenCL supports a
wide range of processors, ranging from embedded and application specific platforms
to more general like x86 multicore CPUs or GPGPUs. With ...

Handbook of Energy-Aware and Green Computing-Two Volume Set

JK Hicks, D Kudithipudi, P Pande, A Ganguly, S Deb…
... 687. Chapter 30. Comparative Study of Runtime Systems for Energy-Aware
High-Performance Computing Chung-Hsing Hsu, Stephen W. Poole Abstract -
Hi-Res PDF (1137 KB) - PDF w/links (1203 KB)xml. 709. Chapter 31. ...

Tuesday, March 22, 2016

Research Readings in Cloud Computing - runtime systems

The VINEYARD Approach: Versatile, Integrated, Accelerator-Based, Heterogeneous Data Centres

C Kachris, D Soudris, G Gaydadjiev, HN Nguyen… - Applied Reconfigurable …, 2016
... with the accelerator programming model in a hybrid solution is a challenging and rather inflexible
proposition, both due to semantical conflicts (eg differences in memory models) and due to
performance implications, notably contention between runtime systems for shared ...

J2M: a Java to MapReduce translator for cloud computing

B Li, J Zhang, N Yu, Y Pan - The Journal of Supercomputing, 2016
... and rigid. To develop the applications with these runtime systems, the developers
should be proficient in the following three domains: MapReduce model, runtime
system, and system-specific programming lan- guage. Although ...

Introducing a Multithread and Multistage Mechanism for the Global Load Balancing Library of X10

K Yamashita, T Kamada - Journal of Information Processing, 2016
... He was a research associate at Kobe University during 1998–2010, and became a lecturer in
2010. His research interest includes parallel and distributed computations, and the runtime
systems of programming lan- guages. c 2016 Information Processing Society of Japan 424

[PDF] 5 Open Problems

J Peters, J Piater, R Platt, S Srinivasa - Multimodal Manipulation Under Uncertainty, 2016
Page 19. Jan Peters, Justus Piater, Robert Platt, and Siddhartha Srinivasa 17 5 Open
Problems The group identified several open problems: 1. Benchmarks for robotic
manipulation: A key challenge is in developing benchmark ...

The VINEYARD Approach: Versatile, Integrated, Accelerator-Based, Heterogeneous Data Centres

HN Nguyen, DS Nikolopoulos, A Bilas, N Morgan… - Applied Reconfigurable …
... with the accelerator program- ming model in a hybrid solution is a challenging and rather inflexible
proposi- tion, both due to semantical conflicts (eg differences in memory models) and due to
performance implications, notably contention between runtime systems for shared ...

[PDF] Towards Asynchronous Many-Task In Situ Data Analysis Using Legion

P Peba, JC BENNETT, D HOLLMAN, S TREICHLER…
... Sandia is a multi-program laboratory operated by Sandia Corporation, a Lockheed Martin
Company, for the US Department of Energy under contract DE-AC04-94AL85000. and increasingly
complex workflows. Asynchronous many task (AMT) models and runtime systems, cf. ...
 

blogger templates 3 columns | Make Money Online