• Journal of KIISE:Computer Systems and Theory
• /
• v.32 no.5
• /
• pp.196-208
• /
• 2005

In multiprocessor systems, interconnection network design is critical for overall system performance. Popular interconnection networks, which are generally considered, are meshes, rings, and hierarchical rings. In this paper, we propose (')Torus Ring('), which is a modified version of hierarchical ring. Torus Ring has the same complexity as the hierarchical rings, but the only difference is the way it connects the local rings. It has an advantage over the hierarchical rings when the destination of a packet is the neighbor local ring in the reverse direction. Though the average number of hops in Torus Ring is equal to that of the hierarchical rings when assuming the uniform distribution of each transaction, the benefits of the number of hops are expected to be larger because of the spatial locality in the real environment of parallel programming. In the simulation results, latencies in the interconnection network are reduced by up to 19$\%$, and the execution times are reduced by up to 10$\%$.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.47 no.10
• /
• pp.57-66
• /
• 2010

Today the various types of communication and application service are provided by the development of Internet and IP technologies. It is expected to be extended the service domain of two networks through the interconnecting P2PSIP overlay which is highlighted as an important technologies for Internet based services and IMS(IP Multimedia Subsystem) which is the new architecture adopted in the evolution of NGN. Therefore, this paper explains the possible methods of the service expansion with interconnecting P2PSIP overlay and IMS network. Specially this paper suggests the interconnection architectures of P2PSIP overlay and IMS network as subscriber's types and analyzes the traffic analysis model and session set up delay characteristics by simulation model. As a results, this paper shows that the interconnection architecture using gateway AS(Application Server) is the excellent method to interconnect the IMS in case of P2PSIP overlay only subscriber, and that interconnection architecture using I-CSCF is good method to interconnect the IMS in case of P2PSIP overlay and IMS subscriber.

• Journal of the Microelectronics and Packaging Society
• /
• v.6 no.4
• /
• pp.61-71
• /
• 1999

The recent trend toward higher frequencies, miniaturization and lower-cost in wireless communication equipment is demanding high density packaging technologies such flip chip interconnection and multichip module(MCM) as a substitute of conventional plastic package. With analyzing the recently reported research results of the RF flip chip, this paper presents the technical issues and advantages of RF flip chip and suggest the flip chip technologies suitable for the development stage. At first, most of RF flip chips are designed in a coplanar waveguide line instead of microstrip in order to achieve better electrical performance and to avoid the interaction with a substrate. Secondly, eliminating wafer back-side grinding, via formation, and back-side metallization enables the manufacturing cost to be reduced. Finally, the electrical performance of flip chip bonding is much better than that of plastic package and the flip chip interconnection is more suitable for Transmit/Receiver modules at higher frequency. However, the characterization of CPW designed RF flip chip must be thoroughly studied and the Au stud bump bonding shall be suggested at the earlier stage of RF flip chip development.

• Journal of the Microelectronics and Packaging Society
• /
• v.6 no.2
• /
• pp.13-21
• /
• 1999

Flip chip assembly directly on organic boards offers miniaturization of package size as well as reduction in interconnection distances resulting in a high performance and cost-competitive Packaging method. This paper describes the investigation of alternative low cost flip-chip mounting processes using electroless Ni/Au bump and anisotropic conductive adhesives/films as an interconnection material on organic boards such as FR-4. As bumps for flip chip, electroless Ni/Au plating was performed and characterized in mechanical and metallurgical point of view. Effect of annealing on Ni bump characteristics informed that the formation of crystalline nickel with $Ni_3$P precipitation above $300^{\circ}C$ causes an increase of hardness and an increase of the intrinsic stress resulting in a reliability limitation. As an interconnection material, modified ACFs composed of nickel conductive fillers for electrical conductor and non-conductive inorganic fillers for modification of film properties such as coefficient of thermal expansion(CTE) and tensile strength were formulated for improved electrical and mechanical properties of ACF interconnection. The thermal fatigue life of ACA/F flip chip on organic board limited by the thermal expansion mismatch between the chip and the board could be increased by a modified ACA/F. Three ACF materials with different CTE values were prepared and bonded between Si chip and FR-4 board for the thermal strain measurement using moire interferometry. The thermal strain of ACF interconnection layer induced by temperature excursion of $80^{\circ}C$ was decreased with decreasing CTEs of ACF materials.

• The KIPS Transactions:PartA
• /
• v.11A no.1
• /
• pp.49-56
• /
• 2004

The performance of the multiprocessor systems is limited by the several factors. The system performance is affected by the processor speed, memory delay, and interconnection network bandwidth/latency. By the evolution of semiconductor technology, off the shelf microprocessor speed breaks beyond GHz, and the processors can be scalable up to multiprocessor system by connecting through the interconnection networks. In this situation, the system performances are bound by the latencies and the bandwidth of the interconnection networks. SCI, Myrinet, and Gigabit Ethernet are widely adopted as a high-speed interconnection network links for the high performance cluster systems. Performance improvement of the interconnection network can be achieved by the bandwidth extension and the latency minimization. Speed up of the operation clock speed is a simple way to accomplish the bandwidth and latency betterment, while its physical distance makes the difficulties to attain the high frequency clock. Hence the system performance and scalability suffered from the interconnection network limitation. Duplicating the link of the interconnection network is one of the solutions to resolve the bottleneck of the scalable systems. Dual-ring SCI link structure is an example of the interconnection network improvement. In this paper, I propose a network topology and a transaction path algorism, which optimize the latency and the efficiency under the duplicated links. By the simulation results, the proposed structure shows 1.05 to 1.11 times better latency, and exhibits 1.42 to 2.1 times faster execution compared to the dual ring systems.

• ETRI Journal
• /
• v.29 no.3
• /
• pp.405-407
• /
• 2007

A globally adaptive load-balanced routing algorithm for torus interconnection networks is proposed. Unlike previously published algorithms, this algorithm employs a new scheme based on collision detection to handle deadlock, and has higher routing adaptability than previous algorithms. Simulation results show that our algorithm outperforms previous algorithms by 16% on benign traffic patterns, and by 10% to 21% on adversarial traffic patterns.

• Proceedings of the International Microelectronics And Packaging Society Conference
• /
• 2001.07a
• /
• pp.95-99
• /
• 2001

Cu is considered as a promising alternative interconnection material to Al-based interconnection materials in Si-based integrated circuits due to its low resistivity and superior resistance to the electromigration. New humping and UBM material systems for solder flip chip interconnection of Cu pads were investigated using electroless-plated copper (E-Cu) and electroless-plated nickel (E-Ni) plating methods as low cost alternatives. Optimally designed E-Ni/E-Cu UBM bilayer material system can be used not only as UBMs for solder bumps but also as bump itself. Electroless-plated E-Ni/E-Cu bumps assembled using anisotropic conductive adhesives on an organic substrate is successfully demonstrated and characterized in this study

• Proceedings of the IEEK Conference
• /
• 1998.10a
• /
• pp.467-470
• /
• 1998

In this paper, we propose a matrix star graph which improves the network cost of the well-known star grah as an interconnection network. We analyze its characteristics in terms of the network parameters, such as degree, scalability, routing, and diameter. The proposed matrix star graph MS2,n has the half degrees of a star graph S2n with the same number of nodes and is an interconnection network with the properties of node symmetry, maximum fault tolerance, and recursive structure. In network cost, a matrix star graph MS2,n and a star graph S2n are about 3.5n2 and 6n2 respectively which means that the former has a better value by a certain constant than the latter has.

• Proceedings of the Korea Electromagnetic Engineering Society Conference
• /
• 2005.11a
• /
• pp.111-114
• /
• 2005

The isolation performance of the S-band single-pole single-throw (SPST) monolithic microwave integrated circuit (MMIC) switch with two different RF-interconnection approaches, microstrip and grounded coplanar waveguide (GCPW) lines, are investigated. On-to-off isolation is improved by 5.8 dB with the GCPW design compared with the microstrip design and additional improvement of 6.9dB is obtained with the coplanar wire-bond interconnection (CWBI) at 3.4 GHz. The measured insertion loss and third-order inter-modulation distortion (IMD3) are less than 2.43 dB over 2.5 CHz $\sim$ 4 GHz and greater than 64 dBc.

• The Journal of Society for e-Business Studies
• /
• v.8 no.1
• /
• pp.1-14
• /
• 2003

The first edition of ebXML which aims at unimarket was announced at May 1 of 2001. OASIS continues working on the framework of ebXML, and UN/CEFACT does on the contents. In our country, 30 vertical B2B markets are being constructed and most of them adopted ebXML as their main standard. In this situation, we need to make a guideline which can interconnect individual vertical B2B systems. In our study, we propose an architecture for i) Central Registry/Repository for the interconnection of between vertical B2Bs, between ebXML and non-ebXML, and between nations. ii) Information modeling for interconnection. iii) Distributed modeling. We hope our work could be extended by following discussion of academical and industrial researchers.