• ETRI Journal
• /
• v.41 no.5
• /
• pp.670-683
• /
• 2019

In this paper, we analyzed the signal integrity of a system interconnection module for a proposed high-density server. The proposed server integrates several components into a chassis. Therefore, the proposed server can access multiple computing resources. To support the system interconnection, among the highly integrated computing resources, the interconnection module, which is based on Serial RapidIO, has been newly adopted and supports a bandwidth of 800 Gbps while routing 160 differential signal traces. The module was designed for two different stack-up types on a printed circuit board. Each module was designed into 12- (version 1) and 14-layer (version 2) versions with thicknesses of 1.5T and 1.8T, respectively. Version 1 has a structure with two consecutive high-speed signal-layers in the middle of two power planes, whereas Version 2 has a single high-speed signal placed only in the space between two power planes. To analyze the signal integrity of the module, we probed the S-parameters, eye-diagrams, and crosstalk voltages. The results show that the high-speed signal integrity of Version 2 has a better quality than Version 1, even if the signal trace length is increased.

• Proceedings of the International Microelectronics And Packaging Society Conference
• /
• 2001.09a
• /
• pp.87-119
• /
• 2001

Fine Pitch Technology will be accelerated among next decade. Buildup Technology is Key Technology for High Density Interconnection. Novel Base Material is critical for High Speed, Area Array Flip Chip Application. Japanese PWB Technology Roadmap will be Published soon.

• Proceedings of the IEEK Conference
• /
• 1998.10a
• /
• pp.289-440
• /
• 1998

This paper describes an interconnection network structure which transports information among processors through a high speed ATM switch. To efficiently use the high speed ATM switch for the message-based multiprocessor, we implemented the cell router that performs multiplexing and demultiplexing of cells from/to processors. In this system, we use the expanded internal cell format including 3bytes for switch routing information. This interconnection network has 3 stage routing strategies: ATM switch routing using switch routing information, cell router routing using a virtual path identifier (VPI) and cell reassembly routing using a virtual channel indentifier (VCI). The interconnection network consists of the NxN folded switch and N cell routers with the M processor interface. Therefore, the maximum number of NxM processors can be interconnected for message communication. This interconnection network using the ATM switch makes a significant improvement in terms of message passing latency and scalability. Additionally, we evaluated the transmission overhead in this interconnection network using ATM switch.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.38 no.8
• /
• pp.1-10
• /
• 2001

This paper describes new line code algorithm, called NIDI(Nibble Inversion mock Inversion) which is well suited for interconnection and transmission technology, The proposed line code which includes only one redundancy bit serves primary features of line code and synchronization patterns for byte or frame synchronization in interconnection, Also, this line code provides in-band signals and speciaI characters.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.48 no.2
• /
• pp.61-66
• /
• 2011

A new via structure on printed circuit board has been proposed for differential signaling in applications of high-speed interconnection. In new structure, the via is physically separated and then divided into two electrically-isolated sections using mechanical drill routing process. These cutted vias are connected respectively to the traces of the differential pair. New via structure makes possible to rout the differential pair using only one via, while conventional via structure needs two vias for interconnection. Because the spacing even in via region keeps almost constant, new via structure can alleviate an impedance discontinuity and then enhance its signal transmission characteristics such as reflection loss and insertion loss. It is expected that new via structure is effective in differential signaling for high-speed interconnection.

• Journal of the Microelectronics and Packaging Society
• /
• v.19 no.4
• /
• pp.13-18
• /
• 2012

By the need for high-speed data transmission in PCB, the studies on the optical PCB has been conducted with optical interconnection and its packaging technology. Particularly, the polymer-based optical interconnection has been extensively studied with the advantages such as cost-effective and ease of process. For high-efficiency and passive alignment, the studies were performed using the 45 degree mirrors, MT connector, and etc. In this work, integrated PLC device and fiber alignment array block was fabricated by using imprint technology to solve the alignment and array problem of optical device and the optical fiber. The fabricated integrated block for optical interconnection of PLC device has achieved higher precision of decreasing the dimensional error of the patterns by optimization of process and its insertion loss has an average value of 4.03dB, lower than criteria specified by international standard. In addition, a optical waveguide with built-in lens has been proposed for high-efficiency and passive alignment. By simulation, it was confirmed that the proposed structure has higher coupling efficiency than conventional no-lens structure and has the broad tolerance for the spatial offset of optical waveguide.

• Proceedings of the IEEK Conference
• /
• 2001.06a
• /
• pp.375-378
• /
• 2001

A 640Gb/s very advanced ATM switching system using 0.25um CMOS VLSI, 40 layer ceramic MCM and 10Gb/s, 8 wavelength 8$\times$8 optical WDM interconnection has been fabricated. To break though the interconnection bottleneck, optical WDM interconnection is used. It has 20Gb/s 8 wavelength 8$\times$8 interconnection capability. It realizes 640Gb/s interconnections within a very small size. Preliminary tests show that 800b1s ATM switch MCM and optical WDM interconnection technologies can be applied to future high speed broadband networks

• 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.

• Proceedings of the IEEK Conference
• /
• 2005.11a
• /
• pp.793-796
• /
• 2005

In this paper, we propose a method which can improve uniformity of a miniaturized electron beam array for inspection of very small pattern with high speed using vertical interconnection. This method enables the individual control of columns so that it can reduce the deviation of beam current, beam size, scan range and so on. The test device that used vertical interconnection method was fabricated by multiple wafer bonding and metal reflow. Two silicon and one glass wafers were bonded and metal interconnection by melting of electroplated AuSn was performed. The contact resistance was under $10{\Omega}$.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.4 no.3
• /
• pp.189-195
• /
• 2004

A 10 Gb/s limiting amplifier IC with the emitter area of $1.5{\times}10{\mu}m^2$ for optical transmission system was designed and fabricated with a AIGaAs/GaAs HBTs technology. In this stud)', we evaluated fine pitch bump using WL-CSP (Wafer Level-Chip Scale Packaging) instead of conventional wire bonding for interconnection. For this we developed WL-CSP process and formed fine pitch solder bump with the $40{\mu}m$ diameter and $100{\mu}m$ pitch on bonding pad. To study the effect of WL-CSP, electrical performance was measured and analyzed in wafer and package module using WL-CSP. In a package module, clear and wide eye diagram openings were observed and the riselfall times were about 100ps, and the output" oltage swing was limited to $600mV_{p-p}$ with input voltage ranging from 50 to 500m V. The Small signal gains in wafer and package module were 15.56dB and 14.99dB respectively. It was found that the difference of small signal gain in wafer and package module was less then 0.57dB up to 10GHz and the characteristics of return loss was improved by 5dB in package module. This is due to the short interconnection length by WL-CSP. So, WL-CSP process can be used for millimeter wave GaAs MMIC with the fine pitch pad.