• Journal of KIISE
• /
• v.42 no.11
• /
• pp.1332-1338
• /
• 2015

Recently, growing attention has been paid to multi-GPU rendering to support real-time high-quality rendering at high resolution. In order to attain high performance in real-time multi-GPU rendering, great care needs to be taken to reduce the overhead of data transfer among GPUs and frame composition. This paper presents a novel multi-GPU algorithm that greatly enhances split frame rendering with implicit query-based synchronization. In order to support implicit synchronization in frame composition, we further present a message queue-based scheduling algorithm. We carried out an experiment to evaluate our algorithm, and found that our algorithm improved rendering performance up to 200% more than previously existing algorithms.

• The KIPS Transactions:PartD
• /
• v.11D no.4
• /
• pp.783-798
• /
• 2004

An e-business client application in ubiquitous mobile computing environment may become disconnected from the enterprise server due to broken communication connections caused by the limitation of mobile computing environments(limited battery life of the mobile device, low bandwidth communication, incomplete wireless communication infrastructure, etc). It Is even Possible that mobile client application Intentionally operates in disconnected mode to reduce communication cost and the power consumption of the mobile device. We use “data hoarding” as a means of providing local autonomy to allow transactions to be processed and committed on the mobile host despite of disconnection. The key problem to this approach is the synchronization problem that serialize potentially conflicting updates from disconnected clients on master objects of the server database. In this paper, we present a new transaction synchronizing method that splits a transaction into a set of independent component transactions and give the synchronization priority on each component taking the possibility of use and conflicts in the server into consideration. Synchronization is performed component by component based un synchronization priority. After the Preferred component of a no bile transaction succeeds in synchronization with the server, the mobile transaction can pre-commit at server. A pre-committed transaction's updated value is made visible at server before the final commit of the transaction. The synchronization of the component with low synchronization priority can be delayed in adaption to wireless bandwidth and computing resources. As a result, the availability of important data updated by mobile client is increased and it can maximize the utilization of the limited wireless bandwidth and computing resources.

• Journal of Korean Institute of Industrial Engineers
• /
• v.3 no.1
• /
• pp.49-53
• /
• 1977

The coordinated control of the traffic signals of adjacent intersections can reduce delays, relative number of stops and congestions in the coordinated traffic area. The road capacity can be increased to a certain extend because the stopping and starting of vehicles facing red traffic lights can be avoided in many instances due to the progression established along an artery. However, if traffic centers or leaves the main flow in irregular volumes on the intermediate road section, a coordination of traffic signals is unnecessary and may even be harmful. Therefore, a computer simulation model to simulate and predict the effectiveness of a synchronized traffic signal system in the CBD of Seoul was developed and alternative policy variables, such as cycle time, offsets, phase splits, to be fed into the simulation model had to be generated. This is a report of (1) the development of a heuristic algorithm for the determination of phase splits when there are amber periods specifically reserved for left turns and (2) the computerization of time-space diagramming.

• KEPCO Journal on Electric Power and Energy
• /
• v.8 no.2
• /
• pp.159-179
• /
• 2022

Often a network becomes complex, and multiple entities would get in charge of managing part of the whole network. An example is a utility grid. While the entire grid would go under a single utility company's responsibility, the network is often split into multiple subsections. Subsequently, each subsection would be given as the responsibility area to the corresponding sub-organization in the utility company. The issue of how to make subsystems of adequate size and minimum number of interconnections between subsystems becomes more critical, especially in real-time simulations. Because the computation capability limit of a single computation unit, regardless of whether it is a high-speed conventional CPU core or an FPGA computational engine, it comes with a maximum limit that can be completed within a given amount of execution time. The issue becomes worsened in real time simulation, in which the computation needs to be in precise synchronization with the real-world clock. When the subject of the computation allows for a longer execution time, i.e., a larger time step size, a larger portion of the network can be put on a computation unit. This translates into a larger margin of the difference between the worst and the best. In other words, even though the worst (or the largest) computational burden is orders of magnitude larger than the best (or the smallest) computational burden, all the necessary computation can still be completed within the given amount of time. However, the requirement of real-time makes the margin much smaller. In other words, the difference between the worst and the best should be as small as possible in order to ensure the even distribution of the computational load. Besides, data exchange/communication is essential in parallel computation, affecting the overall performance. However, the exchange of data takes time. Therefore, the corresponding consideration needs to be with the computational load distribution among multiple calculation units. If it turns out in a satisfactory way, such distribution will raise the possibility of completing the necessary computation in a given amount of time, which might come down in the level of microsecond order. This paper presents an effective way to split a given electrical network, according to multiple criteria, for the purpose of distributing the entire computational load into a set of even (or close to even) sized computational loads. Based on the proposed system splitting method, heavy computation burdens of large-scale electrical networks can be distributed to multiple calculation units, such as an RTDS real time simulator, achieving either more efficient usage of the calculation units, a reduction of the necessary size of the simulation time step, or both.

• Smart Structures and Systems
• /
• v.18 no.6
• /
• pp.1169-1188
• /
• 2016

Real-time substructuring techniques are currently an advanced experimental method for testing large size specimens in the laboratory. In dynamic substructuring, the whole tested system is split into two linked parts, the part of particular interest or nonlinearity, which is tested physically, and the remanding part which is tested numerically. To achieve near-perfect synchronization of the interface response between the physical specimen and the numerical model, a good controller is needed to compensate for transfer system dynamics, nonlinearities, uncertainties and time-varying parameters within the physical substructures. This paper presents the substructuring approach and control performance of the linear and the adaptive controllers for testing the dynamic characteristics of soil-structure-interaction system (SSI). This is difficult to emulate as an entire system in the laboratory because of the size and power supply limitations of the experimental facilities. A modified linear substructuring controller (MLSC) is proposed to replace the linear substructuring controller (LSC).The MLSC doesn't require the accurate mathematical model of the physical structure that is required by the LSC. The effects of parameter identification errors of physical structure and the shaking table on the control performance of the MLSC are analysed. An adaptive controller was designed to compensate for the errors from the simplification of the physical model in the MLSC, and from parameter identification errors. Comparative simulation and experimental tests were then performed to evaluate the performance of the MLSC and the adaptive controller.