• KIPS Transactions on Computer and Communication Systems
• /
• v.8 no.5
• /
• pp.111-118
• /
• 2019

Recent graphic processing units (GPUs) provide high throughput by using powerful hardware resources. However, massive memory accesses cause GPU performance degradation due to cache inefficiency. Therefore, the performance of GPU can be improved by reducing thread parallelism when cache suffers memory contention. In this paper, we propose a dynamic warp scheduler which controls thread parallelism according to degree of cache contention. Usually, the greedy then oldest (GTO) policy for issuing warp shows lower parallelism than loose round robin (LRR) policy. Therefore, the proposed warp scheduler employs the LRR warp scheduling policy when Miss Status Holding Register(MSHR) utilization is low. On the other hand, the GTO policy is employed in order to reduce thread parallelism when MSHRs utilization is high. Our proposed technique shows better performance compared with LRR and GTO policy since it selects efficient scheduling policy dynamically. According to our experimental results, our proposed technique provides IPC improvement by 12.8% and 3.5% over LRR and GTO on average, respectively.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.45 no.6
• /
• pp.978-985
• /
• 2021

This study aims to propose a prediction model for the drape coefficient using artificial neural networks and to analyze the nonlinear relationship between the drape properties and physical properties of fabrics. The study validates the significance of each factor affecting the fabric drape through multiple linear regression analysis with a sample size of 573. The analysis constructs a model with an adjusted R² of 77.6%. Seven main factors affect the drape coefficient: Grammage, extruded length values for warp and weft (m_warp, m_weft), coefficients of quadratic terms in the tensile-force quadratic graph in the warp, weft, and bias directions (c_warp, c_weft, c_bias), and force required for 1% tension in the warp direction (f_warp). Finally, an artificial neural network was created using seven selected factors. The performance was examined by increasing the number of hidden neurons, and the most suitable number of hidden neurons was found to be 8. The mean squared error was .052, and the correlation coefficient was .863, confirming a satisfactory model. The developed artificial neural network model can be used for engineering and high-quality clothing design. It is expected to provide essential data for clothing appearance, such as the fabric drape.

• The Journal of the Korea Contents Association
• /
• v.13 no.3
• /
• pp.9-21
• /
• 2013

Due to increased computing power and flexibility of GPU, recent GPUs execute general purpose parallel applications as well as graphics applications. Programmers can use GPGPU by using the APIs from GPU vendors. Unfortunately, computational resources of GPU are not fully utilized when executing general purpose applications because of frequent branch instructions. To handle the branch problem, several warp formations have been proposed. Intuitively, we expect that the warp formations providing higher computational resource utilization show higher performance. Contrary to our expectations, according to simulation results, the performance of the warp formation providing better utilization is lower than that of the warp formation providing worse utilization. This is because warp formation providing high utilization causes serious memory bottleneck due to increased memory request. Therefore, warp formation providing high computation utilization cannot guarantee high performance without proper hardware resources. For this reason, we will analyze the correlation between hardware configuration and warp formation. Our simulation results present the guideline to solve the underutilization problem due to branch instructions when designing recent GPU.

• KIPS Transactions on Computer and Communication Systems
• /
• v.6 no.5
• /
• pp.219-230
• /
• 2017

General-Purpose Graphics Processing Units (GPGPUs) build massively parallel architecture and apply multithreading technology to explore parallelism. By using programming models like CUDA, and OpenCL, GPGPUs are becoming the best in exploiting plentiful thread-level parallelism caused by parallel applications. Unfortunately, modern GPGPU cannot efficiently utilize its available hardware resources for numerous general-purpose applications. One of the primary reasons is the inefficiency of existing warp/thread block schedulers in hiding long latency instructions, resulting in lost opportunity to improve the performance. This paper studies the effects of hardware thread scheduling policy on GPGPU performance. We propose a novel warp scheduling policy that can alleviate the drawbacks of the traditional round-robin policy. The proposed warp scheduler first classifies the warps of a thread block into two groups, warps with long latency and warps with short latency and then schedules the warps with long latency before the warps with short latency. Furthermore, to support the proposed warp scheduler, we also propose a supplemental technique that can dynamically reduce the number of streaming multiprocessors to which will be assigned thread blocks when encountering a high contention degree at the memory and interconnection network. Based on our experiments on a 15-streaming multiprocessor GPGPU platform, the proposed warp scheduling policy provides an average IPC improvement of 7.5% over the baseline round-robin warp scheduling policy. This paper also shows that the GPGPU performance can be improved by approximately 8.9% on average when the two proposed techniques are combined.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.24 no.2
• /
• pp.78-82
• /
• 1988

Near sea trawlers of Korea sometimes catch pelagic fishes like file fish by using midwater trawl gear even though usually catch bottom fish. It is reasonable to use the specific otter board as well as specific net in bottom trawling and in midwater trawling respectively. But, the trawlers are so small ranging 100 to 120GT, 700 to 100ps that it is very complicated to use different otter board for bottom trawling and for midwater trawling. The otter board for bottom trawling. is also used for the midwater trawling without any change even though the net is changed into the specific one. Although the otter board in the midwater trawling should be lighter than that for bottom trawling, to use otter board for bottom trawling directly for the midwater trawling without any change makes the net easily touch the sea bed and also make the horizontal opening of the otter boards be limited owing to the length of warp in the southern sea of Korea, main fishing ground of midwater trawling, which is 100m or so in depth. That is why the otter board for the midwater trawling should be made lighter than that in the bottom trawling, even if temporary. The authors carried out an experiment to achieve this purpose by attaching a large styropol float on the top of the otter board. In this experiment, underwater weight of the otter board was 630kg and buoyancy of the float was 510kg. To determine the depth and horizontal opening of the otter board, two fish finder was used. A transmitter of 50KHz fish finder was set downward through the shoe plate of otter board to determine the elevation of otter board from the sea bed, and a transmitter of 200KHz fish finder was set sideways on the starboard otter board to be able to detect the distance between otter boards. The obtained results can be summarized as follows: 1. The actual towing speed in the experiment varied 1.1 to 1.8 m/sec. 2. The depth of otter board was within 41 to 25m with float on the top and 45 to 26m without float in case of the warp length 100m, whereas the depth 68-44m with float and 74-46m without float in case of the warp length 150m. This fact means that the depth with float was 9-4% shallower than that without float. 3. The horizontal opening between otter boards was within 34-41m with float and 30-38m without float in case of the warp length 100m, whereas the opening was 44-50m with float and 37-46m without float in case of the warp length 150m. This fact means the opening with float was 10% greater than that without float in case of the warp length 100m, and 15% greater in case of the warp length 150m. 4. The horizontal opening between wing tips by using the otter board with float was 1m greater than by without float in case of the warp length 100m, whereas the opening by with float was 2m greater than by without float in case of warp length 150m. From this fact, it can be estimated that the effective opening area of the net mouth by using the otter board with float could be made 10% greater than by without float in case of warp length 100m, whereas the area with float 20% greater than by without float in case of warp length 150m.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.36 no.4
• /
• pp.281-286
• /
• 2000

To analyze the shape of the net mouth of bottom trawl which is composed with 6 seams net, the field experiment was carried out on the sea near Kokunsan Is, Western sea of Korea. The distance of otter board, net height, trawl speed and resistance of the fishing gear were respectively measured according to the change of warp length and towing speed. The results obtained are summarized as follows : 1. The spreading distance of the otter board has been increased straightly according to the increment of towing speed and warp length. The rate of increase by the warp length has been greatly higher than the rate of increase by the towing speed. The total variation of the spreading distance was 57.0-82.8m, and it was occupied 43-62% of the hand rope, net pendent and the length of nets. 2. The height of net mouth has been decreased straightly according to the increment of towing speed and warp length. The rate of decrease by the towing speed has been greatly higher than the decrease rate of the warp length. The total variation of the net height was 3.1-4.0m. 3. When the distance of wing tip is increased, the height of net mouth is decreased, but the ratio of the decreasing rate of the height of net mouth for the increasing rate of the distance of wing tip was gradually low according to the increment of warp length. 4. The ratio of the distance of both wing tip for the height of net mouth has been increased gradually according to the increment of towing speed and warp length, and the total variation of the ratio was 4.17-7.81 times.

• Fashion & Textile Research Journal
• /
• v.7 no.6
• /
• pp.673-679
• /
• 2005

This study surveys the fabric surface properties such as mean value of the coefficient of friction(MIU), mean deviation of the coefficient of friction(MMD) and mean deviation of surface roughness(SMD) due to warp and weft tension differences using KES-FB system. For this purpose, fabric is designed as 5 harness Satin weave using 150d/48f warp and 200d/384f weft polyester filaments, and is woven by Omega$^{(R)}$ rapier loom by Textec Co.Ltd and Vamatex-P1001ES$^{(R)}$ rapier loom by Vamatex Co.Ltd respectively. These grey fabrics are processed on the same dyeing and finishing processes. The fabric surface properties according to the weaving looms are analysed with warp and weft weaving tensions. And also surveyed the difference of fabric surface properties according to the fabric positions such as center and each edge of fabrics for the sensitive garment. Fabric thickness was also measured and discussed according to the fabric positions such as center and each of fabrics with two looms weaving tensons.

• Textile Coloration and Finishing
• /
• v.12 no.2
• /
• pp.121-128
• /
• 2000

This study discussed the mechanical properties such as bending and shear of polyester fabric treated with a several weight reduction machine. With the increase in the rate of weight loss, the bending rigidity of the warp and weft of treated fabric decreased regardless of the weight reduction machine. At 6.5% weight loss, the bending rigidity of warp and weft yarn decreased to $0.035\;gfcm^2/cm$ and $0.017\;gfcm^2/cm$, respectively, and these values show 54% and 94% of their untreated warp and weft. At same rate of the weight loss, the bending rigidity of polyester fabric treated with C.D.R slightly higher than that of the tank type or liquor-flow type. On the other hand, below 6.5% weight loss, the shear rigidity of the warp and weft of the treated fabric rapidly decreased. But with the increase in the above 6.5% rate of weight loss, the decreasing tendency of the shear rigidity declined. At same rate of the weight loss, the shear rigidity of fabric treated with tank type nearly equal to the that of the liquor flow type. But at same rate of the weight loss, the shear rigidity of the fabric treated with C.D.R type higher than that of the tank or liquor-flow type.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.19 no.3
• /
• pp.543-552
• /
• 1994

Dynamic Time Warp(DTW) needs for iterative calculations and the design of PE cell suitable for the operations is very important. Accordingly, this paper aims at real time recognition design enables large dictionary hardware realization using DTW algorithm. The DTW PE cell separated into three large blocks. "MIN" is the one block for counting accumulated minimum distance. "ADD" block calculates these minimum distances, and "ABS" seeks for the absolute values to the total sum of local distances. Circuit design and verification about the three block have been accomplished, and performed layout '||'&'||' DRC(design rule check) using 1.2 m CMOS N-Well rule base.CMOS N-Well rule base.

• Journal of Fashion Business
• /
• v.23 no.6
• /
• pp.27-36
• /
• 2019

This study aimed to analyze the color differences by warp direction of textile mosaics by focusing on two representative textile wallcovering types, woven and string. Mosaics made of string can be expressed as having three-dimensionality based on color differences resulting from the warp direction of the string. String wallcoverings, unlike woven or non-woven wallcoverings, only have vertically oriented warp lamination on the backing paper without weft, and therefore, the reflection and backing color can be expressed differently depending on the angle of the mosaic. In this study, two identical wallcoverings were manufactured using the same materials but using different textile types, woven and string. The wallcoverings underwent die-cutting by each angle and were deployed in cube form. The analysis was based on ISO 5631-1:2015. The color difference between the two wallcoverings, woven and string, was shown as ΔE* 9.29. Based on the standard deviation of the color difference for each mosaic angle, woven ranged from ΔE* 0.09 to 0.94 and string ranged from ΔE* 1.92 to 3.74, showing a larger color difference. Thus, using the color differences of string to create a mosaic wallcovering improved dimensionality.