This paper describes a tool, LAPSE (Large Application parallelsimulation Environment), that allows one to use a small number of parallel processors to simulate the behavior of a message-passing code running on a larg...
详细信息
ISBN:
(纸本)1565550277
This paper describes a tool, LAPSE (Large Application parallelsimulation Environment), that allows one to use a small number of parallel processors to simulate the behavior of a message-passing code running on a large number of processors, for the purposes of scalability studies and performance tuning. LAPSE is implemented on the Intel Paragon, and has achieved small slowdowns (relative to native code) and high speed-ups on large problems.
Generalized proximity detection for moving objects in a logically correct parallel discrete-event simulation is an interesting and fundamentally challenging problem. Determining who can see whom in a manner that is fu...
详细信息
ISBN:
(纸本)1565550277
Generalized proximity detection for moving objects in a logically correct parallel discrete-event simulation is an interesting and fundamentally challenging problem. Determining who can see whom in a manner that is fully scalable in terms of CPU usage, number of messages, and memory requirements is highly non-trivial. A new scalable approach has been developed to solve this problem. This algorithm, called The Distribution List, has been designed and tested using the object-oriented Synchronous parallel Environment for Emulation and Discrete-Event simulation (SPEEDES) operating system. Preliminary results show that The Distribution List algorithm achieves excellent parallel performance.
IVD, Interactive Visualization Debugger, is intended to provide on-line and integrated mechanisms for debugging, performance analysis, and data visualization for message-passing parallel applications. The current IVD ...
详细信息
ISBN:
(纸本)0818652926
IVD, Interactive Visualization Debugger, is intended to provide on-line and integrated mechanisms for debugging, performance analysis, and data visualization for message-passing parallel applications. The current IVD includes: (1) ESP, a mechanism to multicast window-based commands from a single control window to some subset of existing debuggers/visualizers on various processes;(2) program replay to reproduce program runs deterministically to enable cyclic debugging;and (3) ad hoc data visualization of distributed arrays using existing visualizers.
Fast, efficient parallel algorithms are presented for discrete event simulations of dynamic channel assignment schemes for wireless cellular communication networks. The driving events are call arrivals and departures,...
详细信息
ISBN:
(纸本)1565550277
Fast, efficient parallel algorithms are presented for discrete event simulations of dynamic channel assignment schemes for wireless cellular communication networks. The driving events are call arrivals and departures, in continuous time, to cells geographically distributed across the service area. A dynamic channel assignment scheme decides which call arrivals to accept, and which channels to allocate to the accepted calls, attempting to minimize call blocking while ensuring co-channel interference is tolerably low. Specifically, the scheme ensures that the same channel is used concurrently at different cells only if the pairwise distances between those cells are sufficiently large. Much of the complexity of the system comes from ensuring this separation. The network is modeled as a system of interacting continuous time automata, each corresponding to a cell. To simulate the model, we use conservative methods;i.e., methods in which no errors occur in the course of the simulation and so no rollback or relaxation is needed. Implemented on a 16K processor MasPar MP-1, an elegant and simple technique provides speedups of about I5x over an optimized serial simulation running on a high speed workstation. A drawback of this technique, typical of conservative methods, is that processor utilization is rather low. To overcome this, we developed new methods that exploit slackness in event dependencies over short intervals of time, thereby raising the utilization to above 50% and the speedup over the optimized serial code to about 120x with respect to the workstation simulation.
The performance risks of parallel software development have been shown to cause large escalations in project costs. This paper explains how the VPB performance tools developed by CRL can be used to control such risks....
详细信息
Many real-time applications require more than one processor to achieve the required processing throughput. parallel processing requires the synchronisation of processors, parallel task partitioning and the accessing o...
详细信息
In this paper we make an analogy between the time that storage must be maintained in a optimistic simulation and the blocking time in a conservative simulation. By exploring this analogy, we design two new Global Virt...
ISBN:
(纸本)1565550277
In this paper we make an analogy between the time that storage must be maintained in a optimistic simulation and the blocking time in a conservative simulation. By exploring this analogy, we design two new Global Virtual Time (GVT) protocols for Time Warp systems. The first protocol is based on null message clock advancement in conservative approaches. Our main contribution is a new protocol inspired by Misra's circulating marker scheme for deadlock recovery. It is simple enough to be implemented in hardware, takes no overhead in the normal path, can be made to work over non-FIFO links, and its overhead can be dynamically tuned based on computational load.
Among massively parallel systems, Reconfigurable Processor Array (RPA) [1] is one of the computation models that better support parallel algorithms in the areas of numerical analysis[2], graph analysis[3], vision and ...
详细信息
The proceedings contain 53 papers. The special focus in this conference is on parallel Scientific Computing. The topics include: Computational fluid dynamics on workstation clusters in industrial environments;PVM impl...
ISBN:
(纸本)9783540587125
The proceedings contain 53 papers. The special focus in this conference is on parallel Scientific Computing. The topics include: Computational fluid dynamics on workstation clusters in industrial environments;PVM implementations of advection-chemistry modules of air pollution models;a parallel tree N-body code for heterogeneous clusters;parallel computation of the eigenstructure of toeplitz-plus-hankel matrices on multleomputers;computing eigenvalues of sparse matrices on the connection machine;preliminary experiences with extrapolation methods for parallel solution of differential algebraic equations;a transform approach to fast matrix multiplication;quantum mechanical treatment of reactive scattering problems;load-balancing procedure for distributed short- range molecular dynamics;paralleldistributed 2-D short-range molecular dynamics on networked workstations;implementation of parallel lattice gas program on workstations under PVM;constructing numerical software libraries for high-performance computing environments;the CERFACS experience;towards an automatic portation of programs from nCUBE2 to KSR1;a new algorithm for the traveling salesman problem on networked workstations;nonlinear finite element problems on parallel computers;the challenge of portable libraries for high performance machines;two enhancements in a partitioned sparse code;nonlinear dynamics in biochemical control systems;scientific software libraries for scalable architectures;parallel interval methods for perturbed linear systems;limiting factors in high performance computing;programming the convex exemplar series SPP system;parallelization of HIRLAM on MIMD platforms;simulation of fluid dynamical flow problems;real time simulation in a cluster computing environment and parallel computations in molecular dynamics.
暂无评论