• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2008.05a
• /
• pp.641-644
• /
• 2008

This paper describes the GUI implementation of an intelligent embedded system which can be used for a personal information platform and an industrial mobile application system. It shows the architecture and the necessary structure and components of X Window graphic system. The embedded system used in this paper has low power and high performance processor, and has a large memory size with a nand-flash memory device. We configured the linux kernel with a TIT-LCD and touch screen device for the operation of X Window system. And we used GTK+2 for running application softwares on the platform embedded system. The GUI library of GTK+2 is useful for providing the same graphics programming environment with host Linux PC. We have developed in this paper the X Window system and the GUI environment for GTK+2 in a new embedded system, and verified the full operation of X Window system and application softwares using GTK+2. The embedded system with large memory size can be used in X Window application softwares for providing a personal information service with a mobile embedded system.

• Journal of Information Technology Services
• /
• v.12 no.4
• /
• pp.369-380
• /
• 2013

The advent of the state-of-the-art technologies such as cloud computing and big data processing stimulates the provision of various new IT services, which implies that more servers are required to support them. However, the need for more servers will lead to more energy consumption and the efficient use of energy in the computing environment will become more important. The next generation nonvolatile RAM has many desirable features such as byte addressability, low access latency, high density and low energy consumption. There are many approaches to adopt them especially in the area of the file system involving storage devices, but their focus lies on the improvement of system performance, not on energy reduction. This paper suggests a novel approach for energy reduction in which the MRAM-based SSD is utilized as a storage device instead of the hard disk and a downsized page is adopted instead of the 4KB page that is the size of a page in the ordinary file system. The simulation results show that energy efficiency of a new approach is very effective in case of accessing the small number of bytes and is improved up to 128 times better than that of NAND Flash memory.

• Proceedings of the Korean Information Science Society Conference
• /
• 2007.10c
• /
• pp.140-145
• /
• 2007

최근에 에너지의 효율성이 좋고 대용량화가 쉬운 낸드 플래시가 센서 노드를 위한 차세대 저장소로 각광을 받고 있다. 현재 대부분의 센서 노드용 파일 시스템은 노어 플래시 기반으로 개발되어 있으며 낸드 플래시에 적용할 수 있는 파일 시스템은 거의 존재하지 않는다. 대용량 낸드 플래시 메모리의 특성을 고려한 새로운 파월 시스템의 구축이 요구되지만, 센서 노드는 오직 4-10 KByte의 매우 작은 크기의 메모리를 지원하므로 효율성이 뛰어난 파일 시스템을 구축하는 것은 매우 어렵다. 본 논문은 1 Kbit의 매우 작은 크기의 EEPROM을 부착하여 이러한 메모리 한계를 극복하였으며 자원의 효율성, 대용량의 지원 및 신뢰성을 고려한 새로운 파일 시스템의 설계에 대하여 논한다. 위치를 유지해야 하는 데이터의 위치저장을 위하여 EEPROM을 사용하며 장기간 데이터를 수집할 때 페이지의 갱신을 최소화 할 수 있는 로그 리스트 기반의 페이지 처리 방법에 대해 제안한다. 이는 획기적으로 페이지 갱신 횟수를 줄임으로써 에너지를 절약하고 보다 긴 시간동안 데이터의 수집을 용이하게 만들며 센서 노드의 수명을 증가시킨다.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.5 no.4
• /
• pp.225-233
• /
• 2010

In this paper, we try to improve the performance of the demand paging loader suggested to use the demand paging way that is not based on operating system. The demand paging switching strategy used in the existing operating system can know the recently used pages by running multi-processing. Then, based on it, some page switching strategies have been made for the recently used pages or the frequently demanded pages. However, the strategies based on operating system cannot be applied in single processing that is not based on operating system because any context switching never occur on the single processing. So, this paper is trying to suggest the demand paging switching strategies that can be applied in paging loader running in single process. In the Return-Prediction-Algorithm, we saw the improved performance in the program that the function call occurred frequently in a long distance. And then, in the Most-Frequently-Used-Page-Remain-Algorithm, we saw the improved performance in the program that the references frequently occurred for the particular pages. Likewise, it had an enormous effect on keeping the memory reduction performance by the demand paging and reducing the running time delay at the same time.

• Proceedings of the Korean Institute of Surface Engineering Conference
• /
• 2018.06a
• /
• pp.46-46
• /
• 2018

메모리 소자의 수요가 데스크톱 컴퓨터의 정체와 모바일 기기의 폭발적인 증가로 NAND flash 메모리의 고집적화로 이어져서 3차원 집적 기술의 고도화가 중요한 요소가 되고 있다. 1 mm 정도의 얇은 웨이퍼 상에 만들어지는 메모리 소자는 실제 두께는 몇 마이크로미터 되지 않는다. 수직방향으로 여러 장의 웨이퍼를 연결하면 폭 방향으로 이미 거의 한계에 도달해있는 크기 축소(shrinking) 기술에 의지 하지 않고서도 메모리 소자의 용량을 증대 시킬 수 있다. CPU, AP등의 논리 연산 소자의 경우에는 발열 문제로 3D stacking 기술의 구현이 쉽지 않지만 메모리 소자의 경우에는 저 전력화를 통해서 실용화가 시작되었다. 스마트폰, 휴대용 보조 저장 매체(USB memory, SSD)등에 수 십 GB의 용량이 보편적인 현재, FEOL, BEOL 기술을 모두 가지고 있는 국내의 반도체 소자 업체들은 자연스럽게 TSV 기술과 이에 필요한 장비의 개발에 관심을 가지게 되었다. 특히 이 중 TSV용 스퍼터링 장치는 transistor의 main contact 공정에 전 세계 시장의 90% 이상을 점유하고 있는 글로벌 업체의 경우에도 완전히 만족스러운 장비를 공급하지는 못하고 있는 상태여서 연구 개발의 적절한 시기이다. 기본 개념은 일반적인 마그네트론 스퍼터링이 중성 입자를 타겟 표면에서 발생시키는데 이를 다시 추가적인 전력 공급으로 전자 - 중성 충돌로 인한 이온화 과정을 추가하고 여기서 발생된 타겟 이온들을 웨이퍼의 표면에 최대한 수직 방향으로 입사시키려는 노력이 핵심이다. 본 발표에서는 고전력 이온화 스퍼터링 시스템의 자기장 해석, 냉각 효율 해석, 멀티 모듈 회전 자석 음극에 대한 동역학적 분석 결과를 발표한다. 그림1에는 이중 회전 모듈에 대한 다물체 동역학 해석을 Adams s/w package로 해석하기 위하여 작성한 모델이고 그림2는 180도 회전한 서브 모듈의 위상이 음극 냉각에 미치는 효과를 CFD-ACE+로 유동 해석한 결과를 나타내고 있다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.41 no.10
• /
• pp.1167-1175
• /
• 2016

This paper proposes an efficient page-level mapping algorithm that reduces the erase count in the FTL for flash memory systems. By maintaining the weight for each write request in the request buffer, the proposed algorithm estimates the degree of temporal locality for each incoming write request. To exploit temporal locality deliberately for determination of hot request, the degree of temporal locality should be much higher than the reference point determined experimentally. While previous LRU algorithm treats a new write request to have high temporal locality, the proposed algorithm allows write requests that are estimated to have high temporal locality to access hot blocks to store hot data intensively. The pages are more frequently updated in hot blocks than warm blocks. A hot block that has most of invalid pages is always selected as victim block at Garbage Collection, which results in delayed erase operation and in reduced erase count. Experimental results show that erase count is reduced by 9.3% for real I/O workloads, when compared to the previous LRU algorithm.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.11
• /
• pp.219-225
• /
• 2013

Program and erase characteristics of the elliptic gate all around (e-GAA) SONOS cell have been studied as the variation of eccentricity of the channel. An analytic program and erase model for the elliptic GAA SONOS cell is proposed and evaluated. The model shows that the ISPP (incremental-step-pulse programming) property is changed non-linearly as the eccentricity of the e-GAA SONOS cell is increased. It is differently from the well known linear relationship for that of 2D SONOS and even 3D circular SONOS cell with program bias. We can find that the simulation results of ISPP characteristics are in accord with the experimental data.

• Journal of Sensor Science and Technology
• /
• v.20 no.5
• /
• pp.336-342
• /
• 2011

Recently, the LCD manufacturing industries try to maximize the productivity of LCD panels due to sharp increase in the market need of LCD display. Usually, the increase in manufacturing speed induces additional vibration of glasses and manufacturing machineries. This kind of vibration can induce bad effects on the manufacturing accuracy, and even can damage the glasses. The vibration signal of glass itself is very useful to predict the allowable maximum level of manufacturing speed. Therefore, it is necessary to measure the vibration of the glass itself and a very thin vibration measurement system attachable on the glass is needed. Since in some processes the glasses need to go through very thin gaps such as 2.5mm, We develop a glass vibration measurement system with the thickness of 1.3mm. The system measures the glass vibration using MEMS type accelerometers and store vibration data in a Nand-Flash memory. The performance of the develop system has been verified on a real LCD manufacturing line and the accuracy of vibration measurement is comparable with that of an accurate commercial vibration measurement system.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.48 no.7
• /
• pp.57-64
• /
• 2011

To adjust the difference of bandwidth between host interface and NAND flash memory, DRAM is adopted as the buffer management in SSD (Solid-state Disk). In this paper, we propose cost-effective techniques to enhance SSD performance instead of using expensive high bandwidth DRAM. The SSD data can be classified into three groups such as user data, meta data for handling user data, and FEC(Forward Error Correction) parity/ CRC(Cyclic Redundancy Check) for error control. In order to improve the performance by considering the features of each data, we devise a flexible burst control method through monitoring system and a page based FEC parity/CRC application. Experimental results show that proposed methods enhance the SSD performance up to 25.9% with a negligible 0.07% increase in chip size.

• KIISE Transactions on Computing Practices
• /
• v.21 no.1
• /
• pp.52-57
• /
• 2015

In this paper, we analyze the performance of the TRIM commands on various SSDs and, based on our analysis results, we enhance the performance of these TRIM commands in the Ext4 file system. We observed that the performance of the TRIM commands improves as the size of the LBA-range increases, the sector number is aligned and continuous or more LBA-ranges are notified via a single TRIM command. However, although the performance is better when multiple LBA-ranges are informed by a single TRIM command, the Ext4 file system issues a single TRIM command for every LBA-range. In this paper, we modify the Ext4 file system to convey at most 64 LBA-ranges per TRIM command. Evaluations through Filebench show that the performance of file deletion operations is improved by up to 35%.