• The Transactions of the Korea Information Processing Society
• /
• v.7 no.1
• /
• pp.74-87
• /
• 2000

PVM (parallel virtual machine) provides a programming environment that allows a collection of networked workstations to appear as a single parallel computational resource. The performance of parallel applications in this environment depends on the performance of data transfers between tasks. In this paper, we present a new Myrinet-based communication model of PVM that improves PVM communication performance over a high-speed Myrinet LAN. The proposed PVM communication model adopts a communication mechanism that allows any user-level process to directly access the network interface board without going through UDP/IP protocol stacks in the kernel. This mechanism provides faster data transfers between PVM tasks over the Myrinet since it avoids data copy overhead from kernel (user space) to user space (kernel) and reduces communication latency due to network protocol software layers. We implemented EPVM (Enhanced PVM), our updated version of the traditional PVM using UDP/IP, that is based on the proposed communication model over the Myrinet. Performance results show EPVM achieves communication speed-up of one to two over the traditional PVM.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2004.04a
• /
• pp.430-438
• /
• 2004

Among the optimization method, GA (genetic algorithm) is a very powerful searching method enough to compete with design sensitivity analysis method. GA is very easy to apply, since it dose not require any design sensitivity information. However, GA has been computationally not efficient due to huge repetitive computation. In this study, parallel computation is adopted to Improve computational efficiency, Paralleled GA is introduced on a clustered LINUX based personal computer system.

• Proceedings of the KIEE Conference
• /
• 1996.07a
• /
• pp.68-70
• /
• 1996

In this paper, an efficient parallel computation method for solving large sparse systems of linear algebraic equations by using Cholesky's method in the finite element method is studied. The methods of minimizing the number of fill-ins in the factorization process of factorization are investigated for minimizing the amount of memory and computation time. The parallel programming is implemented under the PVM(Parallel Virtual Machine) environment. The method of load-distribution is studied for minimizing the computation time and the communication time.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.5 no.9
• /
• pp.1544-1572
• /
• 2011

Task placement and scheduling are traditionally studied in following aspects: resource utilization, application throughput, application execution latency and starvation, and recently, the studies are more on application scalability and application performance. A methodology for task placement and scheduling centered on tasks based on virtual machines is studied in this paper to improve the performances of systems and dynamic adaptability in applications development and deployment oriented parallel computing. For parallel applications with no real-time constraints, we describe a thought of feature model and make a formal description for four layers of task placement and scheduling. To place the tasks to different layers of virtual computing systems, we take the performances of four layers as the goal function in the model of task placement and scheduling. Furthermore, we take the personal preference, the application scalability for a designer in his (her) development and deployment, as the constraint of this model. The workflow of task placement and scheduling based on virtual machines has been discussed. Then, an algorithm TPVM is designed to work out the optimal scheme of the model, and an algorithm TEVM completes the execution of tasks in four layers. The experiments have been performed to validate the effectiveness of time estimated method and the feasibility and rationality of algorithms. It is seen from the experiments that our algorithms are better than other four algorithms in performance. The results show that the methodology presented in this paper has guiding significance to improve the efficiency of virtual computing systems.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.5B
• /
• pp.899-912
• /
• 2000

PVM provides users with a single image of high performance parallel computing machine by collecting machines distributed over a network. Low communication overhead is essential to effectively run applications on PVM based platforms. In the original PVM, three times of memory copies are required for a PVM task to send a message to a remote task, which results in performance degradation. We propose a zero-copy model using global shared memory that can be accessed by PVM tasks, PVM daemon, and network interface card(NIC). In the scheme, a task packs data into global shared memory, and notify daemon that the data is ready to be sent, then daemon routes the data to a remote task to which it is sent with no virtual data copy overhead. Experimental result reveals that the message round trip time between two machines is reduced significantly in the proposed zero-copy scheme.

• Transactions of Materials Processing
• /
• v.7 no.5
• /
• pp.474-480
• /
• 1998

In the present study a domain decomposition scheme using the substructuring method is developed for the computational efficiency of the finite element analysis of metal forming processes. in order to avoid calculation of an inverse matrix during the substructuring procedure, the modified Cholesky decomposition method is implemented. As obtaining the data independence by the substructuring method the program is easily paralleized using the Parallel Virtual machine(PVM) library on a work-station cluster connected on networks. A numerical example for a simple upsetting is calculated and the speed-up ratio with respect to various number of subdomains and number of processors. The efficiency of the parallel computation is discussed by comparing the results.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.21 no.8
• /
• pp.1024-1033
• /
• 1997

The past several years have witnessed an ever-increasing acceptance and adoption of parallel processing, both for high performance scientific computing as well as for more general purpose applications. Furthermore with increasing needs to perform the complex flow calculations in an efficient manner, the use of the message passing model on distributed networks has emerged as an important alternative to the expensive supercomputers. This work attempts to provide a generic framework to enable the parallelization of all CFD-related works using the master-slave model. This framework consists of (1) input geometry, (2) domain decomposition, (3) grid generation, (4) flow computations, (5) flow visualization, and (6) output display as the sequential components, but performs computations for (2) to (5) in parallel on the workstation clustering. The flow computations are parallized by having multiple copies of the flow-code to solve a PDE on different spatial regions on different processors, while their flow data are exchanged across the region boundaries, and the solution is time-stepped. The Parallel Virtual Machine (PVM) is used for distributed communication in this work.

• Journal of KIISE:Computer Systems and Theory
• /
• v.29 no.7
• /
• pp.377-383
• /
• 2002

The TSP(Traveling Salesman Problem) has been known as NP-complete, there have been various studies to find the near optimal solution. The nearest neighbor heuristic is more simple than the other algorithms which are to find the optimal solution. This paper designs and implements a new distributed nearest neighbor heuristic with bounding function for the TSP using the master/slave model of PVM(Parallel Virtual Machine). Distributed genetic algorithm obtains a near optimal solution and distributed nearest neighbor heuristic finds an optimal solution for the TSP using the near optimal value obtained by distributed genetic algorithm as the initial bounding value. Especially, we get more speedup using a new genetic operator in the genetic algorithm.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 1998.03a
• /
• pp.246-249
• /
• 1998

In the present study, domain decomposition using the substructuring method is developed for the computational efficiency of the finite element analysis of metal forming processes. In order to avoid calculation of an inverse matrix during the substructuring procedure, the modified Cholesky decomposition method is implemented. As obtaining the data independence by the substructuring method, the program is easily parallelized using the Parallel Virtual Machine(PVM) library on a workstation cluster connected on networks. A numerical example for a simple upsetting is calculated and the speed-up ratio with respect to various domain decompositions and number of processors. Comparing the results, it is concluded that the improvement of performance is obtained through the proposed method.

• Journal of Information Technology and Architecture
• /
• v.10 no.1
• /
• pp.1-11
• /
• 2013

Various computing systems are used in naval ships. Since each system has a single purpose and its applications are tightly coupled with the physical machine, applications cannot share physical resources with each other. It is hard to utilize resources efficiently in conventional naval shipboard computing environment. In this paper, we present an integration architecture for virtualized naval shipboard computing systems based on open architecture. Our proposed architecture integrates individual computing resources into one single integrated hardware pool so that the OS and applications are encapsulated as a VM. We consider the issue of varying needs of all applications in a naval ship that have different purposes, priorities and requirements. We also present parallel VM migration algorithm that improves the process time of resource reallocation of given architecture. The evaluation results with the prototype system show that our algorithm performs better than conventional resource reallocation algorithm in process time.