20 Hits in 2.2 sec

Implementing 2DT on a multiprocessor [chapter]

Yosi Ben-Asher, Gudula Rünger, Reinkard Wilhelm, Assaf Schuster
1994 Lecture Notes in Computer Science  
The implementation of a parallel functional language is discussed. 2DTprograms are composed of local SPMD-computations and global transformations of 2-dimensional data structures leading to a coarse grain  ...  The implementation is made up of virtual processes doing the local computations and abstract processors emulating several virtual processes and taking care of communication.  ...  An implementation of 2DT on a distributed multiprocessor network should be efficient and should preserve the degree of parallelism.  ... 
doi:10.1007/3-540-57877-3_8 fatcat:jpqlhyaqwvh7bpdb7xhmkjqveu

Designing fault-tolerant systems using automorphisms

Shantanu Dutt, John P. Hayes
1991 Journal of Parallel and Distributed Computing  
This paper presents a general theory for modeling and designing fault-tolerant multiprocessor systems in a systematic and efficient manner.  ...  The fault-tolerant design and reconfiguration process of a multiprocessor is modeled by graph automorphisms.  ...  The paper concludes with a brief discussion on how the automorphic method can be used to implement local sparing.  ... 
doi:10.1016/0743-7315(91)90129-w fatcat:hxkhkckpn5c75fddmp6stboi2a

DiamondTorre Algorithm for High-Performance Wave Modeling

Vadim Levchenko, Anastasia Perepelkina, Andrey Zakirov
2016 Computation  
All calculations at each time are conducted in a region near the slope on the current DiamondTorre row. This property can be used to implement a so-called "calculation window".  ...  The asynchrony parameter reaches N z·2DT S 2 ShS 2 since all vertices in a horizontal DiamondTorre slice are asynchronous.  ... 
doi:10.3390/computation4030029 fatcat:cscmhs3jwjexpdyc4uqcmzp4q4

Using Node Diagnosability to Determine t-Diagnosability under the Comparison Diagnosis Model

Chieh-Feng Chiang, Jimmy J.M. Tan
2009 IEEE transactions on computers  
Diagnosis is an essential subject for the reliability of a multiprocessor system.  ...  based on this structure.  ...  is at least 2ðt À pÞ þ 1.  ... 
doi:10.1109/tc.2008.158 fatcat:sgseqrjbunhj5igofks2uqfbbu

Energy Optimization of Memory Intensive Parallel workloads [article]

Chhaya Trehan, Hans Vandierendonck, Georgios Karakonstantis, Dimitrios S. Nikolopoulos
2016 arXiv   pre-print
Finally, we suggest a scheduling criteria for energy aware scheduling of memory intensive parallel applications.  ...  The existing theoretical work on energy mini- mization using Global DVFS (Dynamic Voltage and Frequency Scaling), despite being thorough, ignores the energy consumed by the CPU on memory accesses and the  ...  Plotting fm f 1 against the memory overload factor, 2dt a f 1 (Figure 2 in Appendix), one can observe that as 2dt a f 1 changes from 0 to 1, the optimal ratio changes very quickly and attains the mid  ... 
arXiv:1605.04111v1 fatcat:vma7kk2csbfvbdqic4xw3yhspy

The DiamondCandy Algorithm for Maximum Performance Vectorized Cross-Stencil Computation

Anastasia Yurievna Perepelkina, Keldysh Institute of Applied Mathematics, Vadim Dmitrievich Levchenko
2018 Keldysh Institute Preprints  
It has high locality in terms of the operational intensity, SIMD parallelism support, and is easy to implement.  ...  The implementation details are shown to illustrate how both instruction and data levels of parallelism are used for many-core CPU.  ...  The authors that Colfax Intl. for providing access to the computers with Xeon Phi processors, education on its use, helpful tips on code implementation and execution.  ... 
doi:10.20948/prepr-2018-225-e fatcat:56mko4mronhstagigef4ibt7jy

Yang-Mills lattice on CUDA

Richárd Forster, Ágnes Fülöp
2013 Acta Universitatis Sapientiae: Informatica  
The dynamics of the classical lattice Yang-Mills equations are studied on a 3 dimensional regular lattice.  ...  During the solution of this system we keep the total energy on constant values and it satisfies the Gauss law. The physical quantities are desired to be calculated in the thermodynamic limit.  ...  The actual number of threads running on the GPU comes through the term of warps. A warp is a set of 32 threads in a given Streaming Multiprocessor.  ... 
doi:10.2478/ausi-2014-0010 fatcat:mbahpfznabbmvb7j25j3f6vigi

Nonlinear Systems for Image Processing [chapter]

Saverio Morfu, Patrick Marquié, Brice Nofiélé, Dominique Ginhac
2008 Advances in Imaging and Electron Physics  
Outlooks on microelectronic implementation For each nonlinear processing example, we have attempted to propose an electronic implementation using discrete electronic components.  ...  lastly a positive one.  ...  RECALL ON JACOBIAN ELLIPTIC FUNCTION We recall here the properties of Jacobian elliptic functions used in section II.B.  ... 
doi:10.1016/s1076-5670(08)00603-4 fatcat:goczyhs7dzfztkpkv22ce7dsrm


2011 International Journal of Foundations of Computer Science  
We consider a game-theoretic bin packing problem with identical items, and we study the convergence time to a Nash equilibrium. In the model proposed, users choose their strategy simultaneously.  ...  First, this model is closer to practical situations of large scalable distributed systems, where it may be too expensive, or impossible, to implement a central control responsible for keeping one migration  ...  In case (ii), we have 2Dt(b * ) 3n(b) ≤ 1 for each valid bin b.  ... 
doi:10.1142/s0129054111008234 fatcat:rr6n5bhtfva4ve4mz46liyys5a

Solutions of large-scale electromagnetics problems involving dielectric objects with the parallel multilevel fast multipole algorithm

Özgür Ergül
2011 Optical Society of America. Journal A: Optics, Image Science, and Vision  
Efficiency and accuracy of the developed implementation are demonstrated on very large problems involving as many as 100 million unknowns.  ...  For the solution of large-scale problems discretized with millions of unknowns, MLFMA is parallelized on distributed-memory architectures using a rigorous technique, namely, the hierarchical partitioning  ...  Levent Gürel for fruitful discussions and making invaluable suggestions on this work.  ... 
doi:10.1364/josaa.28.002261 pmid:22048293 fatcat:jwyneaey4rc4xp6d3etqkx4n4i

Symmetric Tori connected Torus Network

Faiz Al Faisal, M.M. Hafizur Rahman
2009 2009 12th International Conference on Computers and Information Technology  
A Symmetric Tori connected Torus Network (STTN) is a 2D-torus network of multiple basic modules, in which the basic modules are 2D-torus networks that are hierarchically interconnected for higher-level  ...  STATIC NETWORK PERFORMANCE Although the actual performance of a network depends on many technological and implementation issues, several topological properties and performance metrics can be used to evaluate  ...  The product (diameter × node degree) is a good criterion for measuring the relationship between cost and performance of a multiprocessor system [3] .  ... 
doi:10.1109/iccit.2009.5407144 fatcat:5fglhszewngrjp2rlawqmhli5q

A novel haptics-based interface and sculpting system for physics-based geometric design

F. Dachille IX, H. Qin, A. Kaufman
2001 Computer-Aided Design  
In this paper, we propose a novel and natural haptic interface and present a physics-based geometric modeling approach that facilitates interactive sculpting of spline-based virtual material.  ...  Conventional geometric design techniques based on B-splines and NURBS often require tedious control-point manipulation and/or painstaking constraint speci®cation via unnatural mouse-based computer interfaces  ...  Because haptics entails a great amount of time-critical processing, a high haptic update rate is often achieved either using a dedicated processor or using a multiprocessor machine. Jacobs et al.  ... 
doi:10.1016/s0010-4485(00)00131-7 fatcat:3aerguhotrg4foutcj6ugzw3ae

Distributed selfish bin packing

Flavio K. Miyazawa, Andre L. Vignatti
2009 2009 IEEE International Symposium on Parallel & Distributed Processing  
We consider a game-theoretic bin packing problem with identical items, and we study the convergence time to a Nash equilibrium. In the model proposed, users choose their strategy simultaneously.  ...  We consider the case when users know the load of all bins and a case with less information. We consider two approaches, depending if the system can undo movements that lead to infeasible states.  ...  First, this model is closer to practical situations of large scalable distributed systems, where it may be too expensive, or impossible, to implement a central control responsible for keeping one migration  ... 
doi:10.1109/ipdps.2009.5160881 dblp:conf/ipps/MiyazawaV09 fatcat:l5ngbbgz35du7d3gnqp5cs367a

Dynamic broadcasting in parallel computing

E.A. Varvarigos, D.P. Bertsekas
1995 IEEE Transactions on Parallel and Distributed Systems  
We consider the problem where broadcast requests are generated at random time instants at each node of a multiprocessor network.  ...  For the dynamic scheme we find an upper bound on the average delay required to serve a broadcast request, and we evaluate its stability region.  ...  During subphase 0 the nodes have (at most) one packet and this is the only one they broadcast. Phase 3 is easy to implement since the current subphase 1 is easily known.  ... 
doi:10.1109/71.342123 fatcat:hvl57435fzcgzee6srilln7giq

Autonomous Formation Flying for the PRISMA Mission

Eberhard Gill, Oliver Montenbruck, Simone D'Amico
2007 Journal of Spacecraft and Rockets  
The guidance and control concept is detailed with emphasis on a relative eccentricity and inclination vector separation strategy.  ...  As a result, the onboard navigation system provides absolute and relative orbit information in real time with a position accuracy of 2 and 0.1 m, respectively.  ...  In [12] , one filter is implemented for the absolute navigation based on pseudorange and Doppler measurements, whereas a second filter for relative navigation assumes a Keplerian motion to process single-difference  ... 
doi:10.2514/1.23015 fatcat:ei3w2lir3bgjblf4zw2xts22zm
« Previous Showing results 1 — 15 out of 20 results