• 대한전기학회논문지:전기물성ㆍ응용부문C
• /
• 제50권4호
• /
• pp.176-181
• /
• 2001

As the data rate between chip-to-chip gets high, the skin effect and load of pins deteriorate noise margin. With these, noise disturbances on the bus channel make it difficult for receiver circuits to read the data signal. This paper has proposed a new pre-emphasis driver circuit which achieves wide noise margin by enlarging the signal voltage range during data transition. When data is transferred from a memory chip to a controller, the output boltage of the driver circuit reaches the final values through the intermediate voltage level. The proposed driver supplies more currents applicable to a packet-based memory system, because it needs no additional control signal and realizes very small area. The circuit has been designed in a 0.18 ${\mu}m$ CMOS process, and HSPICE simulation results have shown that the data rate of 1.32 Gbps be achieved. Due to its result, the proposed driver can achieved higher speed than conventional driver by 10%.

• 한국정보과학회논문지:시스템및이론
• /
• 제32권2호
• /
• pp.55-67
• /
• 2005

시스템 수준 설계에서 계산 부분과 통신 부분의 분리는 프로세서의 선택이나 기능 블록의 프로세서에 대한 할당 결과에 관계없이 설계자로 하여금 독립적인 통신 구조의 설계 공간 탐색을 가능하게 해준다. 본 논문은 버스 기반의 온 칩 통신 구조와 메모리 할당의 최적화를 위한 2단계 설계 공간 탐색 방법을 제안한다. 제안된 설계 공간 탐색 방법은 정적 성능 예측 방법을 사용하여 통신 구조에 대한 방대한 설계 공간을 빠르고 효과적으로 줄인다. 이렇게 축소된 통신 구조들의 설계 공간에 대해서는 정확한 성능 예측을 위하여 프로세서들의 메모리 트레이스론 이용한 트레이스 기반 시뮬레이션을 적용한다. 프로세서들의 동시적인 접근에 의한 버스의 충돌은 프로세서간 공유 메모리뿐 아니라 프로세서의 로컬 메모리에서도 기인하므로 메모리 할당 또한 중요하게 다루어져야 하는 부분이다. 제안된 설계 공간 탐색 방법의 효율성은 4-채널 DVR과 OFDM DVB-T용 수신기 내부의 이퀄라이저 부분을 이용하여 검증하였다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2004년도 하계종합학술대회 논문집(2)
• /
• pp.399-402
• /
• 2004

This paper presents a system-on-chip prototype implementing a full integration for a portable digital audio system. The chip is composed of a audio processor block to implements audio decoding and voice compression or decompression software, a system control block including 8-bit MCU core and Memory Management Unit (MMU) a low power 16-bit ${\Sigma}{\Delta}$ CODEC, two DC-to-BC converter, and a flash memory controller. In order to support other audio algorithms except Mask ROM type's fixed codes, a novel 16-bit fixed-point DSP core with the program-download architecture is proposed. Funker, an efficient power management technique such as task-based clock management is implemented to reduce power consumption for portable application. The proposed chip has been fabricated with a 4 metal 0.25um CMOS technology and the chip area is about 7.1 mm ${\times}$ 7.1mm with 100mW power dissipation at 2.5V power supply.

• 정보처리학회논문지:컴퓨터 및 통신 시스템
• /
• 제11권5호
• /
• pp.133-138
• /
• 2022

최근 고성능컴퓨팅, 인공지능 분야에서 GPU 장치 사용이 일반화되고 있지만, GPU 프로그래밍은 여전히 어렵게 여겨진다. 특히 호스트(host) 메모리와 GPU 메모리를 별도로 관리하기 때문에 성능과 편의성 방면에서 연구가 활발히 진행되고 있다. 이에 따라 여려가지 CPU-GPU 메모리 전송 방법들이 연구되고 있다. 한편 CPU와 GPU 및 통합메모리(Unified memory) 등 하나의 실리콘 패키지로 묶는 SoC(System on a Chip) 제품들이 최근에 많이 출시되고 있다. 본 연구는 이러한 통합메모리 장치에서 CPU, GPU 장치간 데이터를 사용하고 전송시 성능관련 비교를 하고자 한다. 기존 CPU내 호스트 메모리와 GPU 메모리가 분리된 환경과는 다른 특징을 보여준다. 여기서는 통합메모리 장치인 NVIDIA SoC칩들과 NVIDIA SMX 기반 V100 GPU 카드에서 CPU-GPU 간 데이터 전송 프로그래밍 기법별로 성능비교를 한다. 성능비교를 위해 워크로드는 HPC 분야의 수치계산에서 자주 사용하는 2차원 행렬 전치 커널이다. 실험을 통해 CPU-GPU 메모리 전송 프로그래밍 방법별 GPU 커널 성능차이, 페이지 잠긴 메모리와 페이지 가능 메모리를 사용했을 경우 전송 성능차이, 전체(Overall) 성능비교, 마지막으로 워크로드 크기별 성능비교를 하였다. 이를 통해 통합메모리칩인 NVIDIA Xavier에서 I/O 캐시일관성 지원을 통해 SoC 칩내 통합메모리에 대한 이점을 극대화 할 수 있음을 확인할 수 있었다.

• 한국멀티미디어학회논문지
• /
• 제23권2호
• /
• pp.293-300
• /
• 2020

Since the communication networks like the Internet collapses at disaster and calamity sites, a maintenance system that can be operated offline is required for the maintenance of various facilities. In this paper, we propose a system that memory tags attached on the facilities may transmit the emergency manual to a smart-phone, and the smart phone displays it off-line. The main issue is to design low energy mode memory tags. This study presents two kinds of methods and analyzes each's energy consumption mode. The first one is to develop memory tags by using one chip, and the next one is to design memory tags by forming multi-modules. Both ways show proper application fields under the low energy mode. This research selects the off-line maintenance system by using one chip design, and proposes the direction of contents for enhancing the effectiveness of the system. And we expect that this memory tags will be valuable for disaster scenes as well as battle fields.

• 전자공학회논문지A
• /
• 제32A권3호
• /
• pp.89-99
• /
• 1995

In this paper, we ecamined the effect that MCM-L technique influencess on the design and fabrication of multichip memory modules in increasing the packing desity of memory capacity and maximizing its electrical characteristics. For that purpose, we examined the effective methods of reducing the area of module layout and the wiring length with the variation of chip allocation and the number of wiring layers. We fabricated a 256K${\times}$8bit SRAM module with eight 32K${\times}$8bit SRAM chips. The routing experiment showed that we could optimize the area of module layout and wiring length by placing chips in a row, arranging module I/O pads parallel to chip I/O pads, and equalizing the number of terminal sides of module I/O's to that of chip I/O's. The routing was optimized when we used three wire layers in case of one sided chip mounting or five wire layers in case of double sided chip mounting. The fabricated modules showed 18.9 cm/cm$^{2}$ in wiring density, 65 % in substrate occupancy efficiency, and module substrate and functionally tested to find out the module working perfectly.

• 대한임베디드공학회논문지
• /
• 제14권2호
• /
• pp.87-96
• /
• 2019

Low power issue is one of the most critical problems in the Internet of Things (IoT), which are powered by battery. To solve this problem, various approaches have been presented so far. In this paper, we propose a method to reduce the power consumption by reducing the numbers of accesses into the flash memory consuming a large amount of power for on-chip software execution. Our approach is based on using cooperative logging structure to distribute the sampling overhead in single sensor node to adjacent nodes in case of rare-event applications. The proposed algorithm to identify event occurrence is newly introduced with negative feedback method by observing difference between past data and recent data coming from the sensor. When an event with need of flash access is determined, the proposed approach only allows access to write the sampled data in flash memory. The proposed event detection algorithm (EDA) result in 30% reduction of power consumption compared to the conventional flash write scheme for all cases of event. The sampled data from the sensor is first traced into the random access memory (RAM), and write access to the flash memory is delayed until the page buffer of the on-chip flash memory controller in the micro controller unit (MCU) is full of the numbers of the traced data, thereby reducing the frequency of accessing flash memory. This technique additionally reduces power consumption by 40% compared to flash-write all data. By sharing the sampling information via LoRa channel, the overhead in sampling data is distributed, to reduce the sampling load on each node, so that the 66% reduction of total power consumption is achieved in several IoT edge nodes by removing the sampling operation of duplicated data.

• 한국음향학회지
• /
• 제15권6호
• /
• pp.16-25
• /
• 1996

본 논문에서는 한글의 유형 분류를 CAM(Content Addressable Memory)을 이용하여 칩으로 설계하였다. 문자 인식의 전 과정을 종전의 소프트웨어에 의해서 손차적으로 처리할 경우, 실시간 처리가 가능한 고속 문자 인식기의 구현에는 어려움이 있다. 따라서, 이들 실시간으로 처리하기 위해서는 파이프라인식 하드웨어로 구현하여 시간적인 병렬성을 갖도록 하는 것이 필요하다. 하드웨어로 용이하게 구현하기 위해서 고속 병렬 매치 기능을 가진 CAM을 이용하였으며, 20개의 참조패턴만으로 유형을 분류하였다. 설계한 회로는 DAZIX의 DLAB을 사용하여 결과를 평가하였으며, 그 결과 자당 $1.6{\mu}S$의 처리속도를 보였다. 또한, C-언어로 구현하여 그 결과를 비교하였다.

• Design & Manufacturing
• /
• 제10권2호
• /
• pp.50-54
• /
• 2016

In the semiconductor industry the final products were checked for several environments before sell the products. The burning test of memory and chip was implemented in reliability for all of parts. The memory and chip were developed to high density memory and high performance chip, so circuit design was also high integrated and the test bed was needed to be thin and fine pitch socket. LGA(Land Grid Array) IC socket with thin wall thickness was designed to satisfy this requirement. The LGA IC socket plastic part was manufacture by injection molding process, it was needed accuracy, stiffness and suit resin with high flowability. In this study, injection molding process analysis was executed for 2 and 4 cavities moldings with runner, gate and sprue. The warpage analysis was also implemented for further gate removal process. Through the analyses the total deformations of the moldings were predicted within maximum 0.05mm deformation. Finally in consideration of these results, 2 and 4 cavities molds were designed and made and tested in injection molding process.

• 대한전기학회논문지:전기기기및에너지변환시스템부문B
• /
• 제51권4호
• /
• pp.180-186
• /
• 2002

This paper describes on the development of power detection system with one-chip microcontroller. The designed system is composed of power detection circuits and analyzing software. The system detects, 3-phases voltage, 3-phases current, external temperature, leakage current and stores in flash memory. AT89C52 was used as CPU and AM29F040B was used as memory to store the data. The analysis saftware was developed to detect the cause of the electrical fire incidents. With a data-compression technology, the data can be stored for the 43.5 days in a normal state, four hours and fifteen minutes in emergency state.