• Journal of KIISE:Computer Systems and Theory
• /
• v.26 no.11
• /
• pp.1305-1317
• /
• 1999

NUMA 구조는 원격 메모리에 대한 접근이 불가피한 구조적 특성 때문에 상호 연결망이 성능을 좌우하는 큰 변수가 된다. 기존에 대중적으로 사용되던 버스는 물리적 확장성 및 대역폭에서 대규모 시스템을 구성하는 데 한계를 보인다. 이를 대체하는 고속의 지점간 링크를 사용한 링 구조는 버스가 가지는 확장성 및 대역폭의 한계라는 단점을 개선하였으나, 많은 클러스터가 연결되는 경우에는 전송 지연시간이 증가하는 문제점을 가지고 있다. 본 논문에서는 스누핑 프로토콜이 적용된 링 구조에서 클러스터 개수 증가에 따른 지연시간 증가의 문제점을 보완하기 위해 계층적 링 구조로의 확장을 제안하고, 이 구조에 효과적인 캐쉬 일관성 프로토콜을 설계하였다. 전역 링과 지역 링을 연결하는 브리지는 캐쉬 프로토콜을 관리하며 이 프로토콜에 의해 지역 링의 부하를 줄일 수 있도록 트랜잭션을 필터링하는 역할도 담당함으로써 시스템의 성능을 향상시킨다. probability-driven 시뮬레이터를 통해 계층적 링 구조가 시스템의 성능 및 링 이용률에 미치는 영향을 알아본다. Abstract Since NUMA architecture has to access remote memory, interconnection network performance determines performance of NUMA architecture. Bus, which has been used as popular interconnection network of NUMA, has a limit to build a large-scale system because of limited physical scalability and bandwidth. Ring interconnection network, composed of high-speed point-to-point link, made up for bus's defects of scalability and bandwidth. But, it also has problem of increasing delay as the number of clusters is increased. In this paper, we propose a hierarchical expansion of snoop-based ring architecture in order to overcome ring's defects of increasing delay. And we also design an efficient cache coherence protocol adopted to this architecture. Bridge, which connects local ring and global ring, maintains cache coherence protocol and does snoop-filtering which reduces local ring and cluster bus utilization. Therefore bridge can improve performance of this system. We analyze effects of hierarchical architecture on the performance of system and utilization of point-to-point links using probability-driven simulator.

• Journal of KIISE:Computer Systems and Theory
• /
• v.29 no.9
• /
• pp.503-513
• /
• 2002

Since CC-NUMA architecture has to access remote memory, the interconnection network determines the performance of the CC-NUMA system. Bus which has been used as a popular interconnection network has many limits in a large-scale system because of the limited physical scalability and bandwidth. The dual ring interconnection network, composed of high-speed point-to-point links, is made to resolve the defects of the bus for the large-scale system. However, it also has a problem, in that the response latency is rapidly increased when many nodes are attached to the snooping based CC-NUMA system with the dual ring. In this paper, we propose a ring architecture with repeater nodes in order to overcome the problem of the dual ring on a snooping based CC-NUMA system, and design a repeater node adapted to this architecture. We will also analyze the effects of proposed architecture on the system performance and the response latency by using a probability-driven simulator.

• Journal of KIISE:Computer Systems and Theory
• /
• v.29 no.2
• /
• pp.105-115
• /
• 2002

Since NUMA architecture has to access remote memory an interconnection network determines the performance of CC-NUMA system Bus which has been used as a popular interconnection network has many limits to build a large-scale system because of the limited physical scalabilty and bandwidth Dual ring interconnection network composed of high speed point-to-point links is made up for resolving the defects of the bus for large-scale system But it also has a problem that the response latency is rapidly increased when many node are attached to snooping based CC-NUMA system with dual ring In this paper we propose a chordal ring architecture in order to overcome the problem of the dual ring on snooping based CC-NUMA system and design and efficient link controller adopted to this architecture. We also analyze the effects of chordal ring architecture on the system performance and the response latency by using probability driven simulator.

• Journal of the Korean Institute of Telematics and Electronics D
• /
• v.36D no.7
• /
• pp.66-79
• /
• 1999

We investigate the effects of high input power on the performance of optical bistable symmetric self-electooptic effect devices (S-SEEDs) using extremely shallow quantum wells (ESQWs). In this study, we consider the four ESQWs SEEDs; anti-reflection (AR)-coated ESQWs S-SEED, back-to-back AR coated ESQWs S-SEED, asymmetric F뮤교-Perot (AFP) ESQWs S-SEED, and back-to-back AFP-ESQWs S-SEED. As the input power increases, device performances such as on/off contrast ratio, on/off reflectivity difference are seriously degraded because of ohmic heating and exciton saturation. On the other hand, switching speed of the device increases up to certain value and then begins to decrease. With reasonable optimization of the input power for the best switching speed operation of the devices in a cascading optical interconnection system, we simulate and analyze the system bit-rate of the various ESQWs S-SEEDs, for a mesa of $5{\times}5{\mu}m^2$ size, changing the namber of quantum wells for the external bias of 0 V and -5V.

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.4
• /
• pp.961-968
• /
• 1996

In the frame of Koreasat Project, it has been identified the task to implement a pilot satellite network module to provide LAN-to-LAn in ground system for satellite links. The pilot network will support an experiment to verify the performances of the considered applications through a satellite.This paper proposes a satellite-LAN interconnecting architecture making full use of satellite benefits and counteracting satellite demerits. The architectureprovides high quality data transmission and high perfrmance for satellite bit errors by using a connection- oriented satellite protocol which can establish multiple logical links between two nodes. As a protocol conversion method, router-type interconnection was selected to guard against problems. Based on this architecture, a satellite LAN interconnecting system has been designed, which includes a 1.8 meter antenna with a 4 watt transceiver, a satellite modem and the developed satellite network interface. The system can support high speed transmission rates of up to 1.544 Mbs and superior network management as well.

• Journal of the Microelectronics and Packaging Society
• /
• v.5 no.2
• /
• pp.21-26
• /
• 1998

대용량, 고속데이터 처리가 요구되는 System 개발은 이들의 복잡하고 고기능의 회 로 구현이 가능하냐에 달려 있고 또한 이들고기능 요구를 가장 잘 만족할수 있는 패키지는 MCM 이라 할 수있다. 시스템의 고속화, 소형화는 회로의 복잡성을 요구하는 있는 이를 패 키지로 구현하는 MCM은 시험성 확보에 심각한 문제점으로 나타나고 있다. 본 논문에서는 고밀도 구조의 MCM 기판에 대한 Interconnetion Line 시험검증을 위한 Flying Prober의 적 용 및 모듈 패키징 공정에 대한 조립성 검증을 위한 BST에 대해 설명한다. 연구에 사용된 MCM 모듈은 MCM-D 공정으로 제작되었으며 31um 신호선폭, 50um Via Hole Dia. 5신호 선층 5절연층 및 455 Net의 기판으로절연층은 Dow chemical의 BCB-4024/4026을 적용하였 다. 조립은 3 ASIC, 24소자 실장 및 2000 Wire Bonding으로 이루어지며 패키지는 방열특성 을 고려한 BGA(491 I /O,50mil pitch)를 개발하여 사용하였다. MCM 기판의미세패턴으로 구성된 Interconnection Line에 대해 Fine Ptich Probing이 가능한 Flying Prober를 사용하 여 평가하였으며 BST를 이용하여 실장소자의 KGD평가 및 능동, 수동소자가 실장된 MCM Package의 조립시험성을 확보할수 있었다.

• Proceedings of the Korea Contents Association Conference
• /
• 2009.05a
• /
• pp.642-647
• /
• 2009

We proposed the method of avionics integrated architecture using high-speed fiber optic bus. Typically, data bus of aircraft consists of electronic and optic data transmission method. Avionics systems are difficult to operate the electronic data transmission method for the high speed data processing, synchronization and interconnection between flight control system and flight management system efficiently. In this paper, it is known to look into the problem of data bus and the advanced trend in avionics systems, and propose the appropriate data bus of the advanced avionics systems.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.37 no.4
• /
• pp.35-41
• /
• 2000

In this paper, the multiplicative algorithm of two polynomals over finite field GF(2$^{m}$ ) is presented. The proposed algorithm permits an efficient realization of the parallel multiplication using iterative arrays. At the same time, it permits high-speed operation. This multiplier is consisted of three operation unit: multiplicative operation unit, the modular operation unit, the primitive irreducible operation unit. The multiplicative operation unit is composed of AND gate, X-OR gate and multiplexer. The modular operation unit is constructed by AND gate, X-OR gate. Also, an efficient pipeline form of the proposed multiplication scheme is introduced. All multipliers obtained have low circuit complexity permitting high-speed operation and interconnection of the cells are regular, well-suited for VLSI realization.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.2 no.3
• /
• pp.439-446
• /
• 1998

This paper presents design of SD adder and PD adder on Multi-Valued Logic. For implementing of Multi-valued logic circuits we use Current-mode CMOS circuits and also use Voltage-mode CMOS circuits partially. The proposed arithmetic circuits was estimated by SPICE simulation. At the SD(Signed-Digit) number presentation applying Multi-Valued logic the carry propagation is always limited to one position to the left this number presentation allows fast parallel operation. The addition method that add M operands using PD( positive digit number) is effective not only for the realization of the high-speed compact arithmetic circuit, but also for the reduction of the interconnection in the VLSI processor. therefor, if we use PD number representation, the high speed processor can be implementation.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.39 no.2
• /
• pp.91-103
• /
• 2002

For massive FPGA based emulator, the interconnection architecture and the transmission method of signals between FPGA's are important elements which decide speed of emulation and extendability of emulator. Existing FPGA-based emulation system is faced the problems of which the emulation speed getting slow drastically as the complexity of circuit increases. In this paper, we proposed a new innovative emulation architecture that has high resource usage rate and makes the fast emulation Possible. The emulator with very unique hierarchical ring topology Presented here has merits to overcome FPGA pin limitation by connecting each FPGA into a set of pipelined rings, and it also makes emulation speed at the tens of MHz at least even at system level where the verification complexity can easily exceed the verification capability of designers.