• JSTS:Journal of Semiconductor Technology and Science
• /
• v.16 no.6
• /
• pp.808-816
• /
• 2016

As the number of CPU/GPU cores and IPs in SOC increases and applications require explosive memory bandwidth, simultaneously achieving good throughput and fairness in the memory system among interfering applications is very challenging. Recent works proposed priority-based thread scheduling and channel partitioning to improve throughput and fairness. However, combining these different approaches leads to performance and fairness degradation. In this paper, we analyze the problems incurred when combining priority-based scheduling and channel partitioning and propose dynamic priority thread scheduling and adaptive channel partitioning method. In addition, we propose dynamic address mapping to further optimize the proposed scheme. Combining proposed methods could enhance weighted speedup and fairness for memory intensive applications by 4.2% and 10.2% over TCM or by 19.7% and 19.9% over FR-FCFS on average whereas the proposed scheme requires space less than TCM by 8%.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.31 no.6
• /
• pp.8-16
• /
• 2003

Aerodynamic characteristics over Micro Aerial Vehicle(MAV) in low Reynolds number regime are numerically studied using 3-D unsteady, incompressible Navier-Stokes flow solver with single partitioning method for multi-block grid. For more efficient computation of unsteady flows, this flow solver is parallel-implemented with MPl(Message Passing Interface) programming method. Firstly, MAV wing with not complex geometry is considered and then, we analyze aerodynamic characteristics over full MAV configuration varying the angle of attack. Present computational results show a better agreement with the experimental data by MACDL(Micro Aerodynamic Control and Design Lab.), Seoul National University. We can also find the conceptually designed MAV by MACDL has the static stability.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.7
• /
• pp.799-803
• /
• 2010

The KDB-tree is a traditional indexing scheme for retrieving multidimensional data. Much research for KDB-tree family frequently addresses the low storage utilization and insufficient retrieval performance as their two bottlenecks. The bottlenecks occur due to a number of unnecessary splits caused by data insertion orders and data skewness. In this paper, we propose a novel index structure, called as $KDB_{CS}^+$-tree, to process skewed data efficiently and improve the retrieval performance. The $KDB_{CS}^+$-tree increases the number of fan-outs by exploiting bit-vectors for representing splitting information and pointer elimination. It also improves the storage utilization by representing entries as a hierarchical structure in each internal node.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.48 no.8
• /
• pp.565-571
• /
• 2020

In this paper, precision and efficient nonlinear structural analysis for the aircraft wing-box model is developed. Herein, nonlinear shell element based on the co-rotational (CR) formulation is implemented. Then, parallel computing algorithm, the element-based partitioning technique is developed to accelerate the computational efficiency of the nonlinear structural analysis. Finally, computational performance, i.e., accuracy and efficiency, of the proposed analysis is evaluated by comparing with that of the existing commercial software.

• KIISE Transactions on Computing Practices
• /
• v.20 no.9
• /
• pp.519-524
• /
• 2014

The operating systems in the system architecture, which is integrated several applications and modular electronic devices in the same computing device, demand partitioning technology for safety. Thus, operation system requires partition scheduler for partition scheduling. When we design partition scheduler in embedded system, which has small memory and low performance, such as space system, we must consider not only performance but also memory. In this paper, we introduces a linux-based memory efficient partition scheduler using partition bitmap. This partition scheduler demands small memory space and produce low partition switching overhead. The prototype was executed on a LEON4 processor, which is the Next Generation Multicore Processor (NGMP) in the space sector. In evaluation, this prototype shows accuracy, additional memory space and low partition switching overhead.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.11
• /
• pp.20-26
• /
• 1994

In this paper, we propose an algorithm on partitioning a network into several subsets where the cost of a net which connects nodes in k subsets is given as k(k-1)/2 indicating the typical pattern of complete graphs. This problem is one of generalizations for multiple-way partitioning proposed by Sanchis. $^{[5]}$ Its solution can be applied to resource allocation problem in distributed systems. The proposed algorithm expanded the algorithm of Fiduccia and Mattheyses$^{[3]}$ to handle the multiple-way partitioning simultaneously. It has time and space complexity linear to the size of the network. To evaluate the performance of the proposed algorithm, we implemented also a traditional cluster growth method which groups connected nodes for nets, and compared experimental results with those of the proposed algorithm. The proposed algorithm shows some enhancement made.

• KIISE Transactions on Computing Practices
• /
• v.20 no.9
• /
• pp.507-512
• /
• 2014

Recently, Distributed computing processing begins using both CPU(Central processing unit) and GPU(Graphic processing unit) to improve the performance to overcome darksilicon problem which cannot use all of the transistors because of the electric power limitation. There is an integrated graphics processor that CPU and GPU share memory and Last level cache(LLC). But, There is no LLC access rules between CPU and GPU, so if GPU and CPU processes run together at the same time, performance of both processes gets worse because of the contention on the LLC. This Paper gives evidence to prove the need of the Cache Partitioning and is mentioned about the cache partitioning design using page coloring to allocate the L3 Cache space only for the GPU process to guarantee GPU process performance.

• Korean Management Science Review
• /
• v.23 no.3
• /
• pp.13-25
• /
• 2006

To reduce the signaling tost of paging in mobile communication, sequential paging schemes are proposed by partitioning a location area into several paging areas such that each area is paged sequentially. Necessary conditions for the optimal partition of cells with delay bound are examined by considering the mobiles location probability at each cell. The Optimal Cell Partitioning (OCP) is proposed based on the necessary conditions and the fathoming rule which trims off the unnecessary solution space and expedite the search process. Two Heuristics, BSG and BNC are also presented to further increase the computational efficiency in real-world paging scheme for the next generation mobile systems. The effectiveness of the 1)reposed paging schemes is illustrated with computational results. The Heuristic BSG that performs the search in the most promising solution group outperforms the best existing procedure with the 6-69% gain in paging cost in problems with 100 cells.

• Journal of the Korean Society for Railway
• /
• v.2 no.1
• /
• pp.47-55
• /
• 1999

In this paper, an efficient and accurate formulation for the transient analysis of constrained multibody systems is presented. The formulation employs Kane's method along with the null space method. Kane's method reduces the dimension of equations of motion by using partial velocity matrix: it can improve the efficiency of the formulation. Furthermore, the formulation partitions the coefficient matrix of linear and nonlinear equations into several sub-matrices using kinematic uncoupling. This can solve the equations more efficiently. The proposed formulation can be used to perform dynamic analysis of systems which can be partitioned into several sub-systems such as train systems. One numerical example is given to demonstrate the efficiency and accuracy of the formulation, and another numerical example is given to show its application to the train systems.

• Proceedings of the IEEK Conference
• /
• 1999.11a
• /
• pp.1011-1014
• /
• 1999

In this paper, a new algorithm has been presented that helps to preserve diversity as well as to enhance the convergence speed of the evolutionary programming. This algorithm is based on the cell partitioning of search region for preserving the diversity. Until now, the greater part of researches is not concerned about preserving the diversity of individuals in a population but improving convergence speed. Although these evolutions are started from multi-point search at the early phase, but at the end those search points are swarming about a one-point, the strong candidate. These evolutions vary from the original idea in some points such as multi-point search. In most case we want to find the only one point of the best solution not several points in the vicinity of that. That is why the cell partitioning of search region has been used. By restricting the search area of each individual, the diversity of individual in solution space is preserved and the convergence speed is enhanced. The efficiency of the proposed algorithm has been verified through benchmark test functions.