• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.33A no.3
• /
• pp.152-162
• /
• 1996

Diffraction gratings were designed to generate multiple beams for the real-time optical interconnection in free-space using a liquid crsytal television. Multi-level phase gratings consisted of stepped phase value, 0 to 2$\pi$, and generate a 5$\times$5 spot array, and V-shaped spot patterns which have a maximum diffraction efficiency of 78.84%. Simulated annealing technique was used to determine phase states of each pixel having uniform intensity and high diffraction efficiency. Through the optical interconnection experiments, we compared the diffraction characteristics of beam paterns obtained by binary-level, four-level, and eight-level phase gratings.

• Proceedings of the Korean Vacuum Society Conference
• /
• 1992.02a
• /
• pp.81-83
• /
• 1992

• Journal of Information Processing Systems
• /
• v.18 no.1
• /
• pp.159-172
• /
• 2022

The role of the interconnect network, which connects computing nodes to each other, is important in high-performance computing (HPC) systems. In recent years, the peripheral component interconnect express (PCIe) has become a promising interface as an interconnection network for high-performance and cost-effective HPC systems having the features of non-transparent bridge (NTB) technologies. OpenSHMEM is a programming model for distributed shared memory that supports a partitioned global address space (PGAS). Currently, little work has been done to develop the OpenSHMEM library for PCIe-interconnected HPC systems. This paper introduces a prototype implementation of the OpenSHMEM library through a switchless interconnect network using PCIe NTB to provide a PGAS programming model. In particular, multi-interrupt, multi-thread-based data transfer over the OpenSHMEM shared memory model is applied at the implementation level to reduce the latency and increase the throughput of the switchless ring network system. The implemented OpenSHMEM programming model over the PCIe NTB switchless interconnection network provides a feasible, cost-effective HPC system with a PGAS programming model.

• Proceedings of the IEEK Conference
• /
• 2003.07b
• /
• pp.791-794
• /
• 2003

The increasing computational capability of processors is driving the need for high bandwidth links to communicate and store the information that is processed. Such links are often an important part of multi processor interconnection, processor-to-memory interfaces and Serial-network interfaces. This paper describes a 0.11-${\mu}{\textrm}{m}$ CMOS 4 Gbp s/pin 3-Level transceiver using RSL/(Rambus Signaling Logic) for high bandwidth. This system which uses a high-gain windowed integrating receiver with wide common-mode range which was designed in order to improve SNR when operating with the smaller input overdrive of 3-Level. For multi-gigabit/second application, the data rate is limited by Inter-Symbol Interference (ISI) caused by low pass effects of channel, process-limited on-chip clock frequency, and serial link distance. In order to detect the transmited 4Gbps/pin with 3-Level data sucessfully ,the receiver is designed using 3-stage sense amplifier. The proposed transceiver employes multi-level signaling (3-Level Pulse Amplitude Modulation) using clock multi phase, double data rate and Prbs patten generator. The transceiver shows data rate of 3.2 ~ 4.0 Gbps/pin with a 1GHz internal clock.

• Journal of the Microelectronics and Packaging Society
• /
• v.9 no.3
• /
• pp.1-6
• /
• 2002

A low temperature process technology of copper thin films has been developed by a chemical vapor deposition technology for multi-level metallzations in ULSI fabrication. The copper films were deposited on TiN/Si substrates in helium atmosphere with the substrate temperature between $130^{\circ}C$ and $250^{\circ}C$. In order to get more reliable metallizations, effects on the post-annealing treatment to the electrical properties of the copper films have been investigated. The Cu films were annealed at the $5 \times10^{-6}$ Torr vacuum condition and the electrical resistivity and the nano-structures were measured for the Cu films. The electrical resistivity of Cu films shown to be reduced by the post-annealing. The electrical resistivity of 2.0 $\mu \Omega \cdot \textrm{cm}$ was obtained for the sample deposited at the substrate temperature of $180^{\circ}C$ after vacuum annealed at $300^{\circ}C$. The resistivity variations of the films was not exactly matched with the size of the nano-structures of the copper grains, but more depended on the contamination of the copper films.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2006.06a
• /
• pp.303-310
• /
• 2006

The sorting algorithms have been developed to take advantage of distributed computers. But the speedup of parallel sorting algorithms decrease rapidly with increased number of processors due to parallel processing overhead such as context switching time and inter-processor communication cost. In this paper, we propose a parallel sorting method which provides linear speedup of an optimal serial algorithm for a system with a large number of processors. This algorithm may even provide superlinear speedup for a practical system. The algorithm takes advantage of an interconnection network properties and its protocol.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.57 no.3
• /
• pp.416-420
• /
• 2008

This paper proposes a new method to improve signal characteristics at branches frequently met in system-level routing. We also introduce the numerical formula which can estimate the time delay due to branches and the simple design guideline for system-level routing. Finally, we propose the routing method which can eliminate the signal reflection for the case of one driver and two receivers (multi-drop topology).

• The Transactions of the Korean Institute of Electrical Engineers C
• /
• v.51 no.7
• /
• pp.311-315
• /
• 2002

Chemical mechanical polishing (CMP) process was required for the global planarization of inter-metal dielectric(IMD), inter-level dielectric (ILD) layers of multi-layer interconnections. In this paper, we studied the characteristics of polishing pad, which can apply shallow trench isolation (STI)-CMP process for global planarization of multi-level interconnection structure. Also, we investigated the effects of different sets of polishing pad, such as soft and hard pad. As an experimental result, hard pad showed center-fast type, and soft pad showed edge-fast type. Totally, the defect level has shown little difference, however, the counts of scratch was detected less than 2 on JR111 pad. Through the above results, we can select optimum polishing pad, so we can expect the improvements of throughput and device yield.

• Journal of the Korea Society of Computer and Information
• /
• v.16 no.6
• /
• pp.1-10
• /
• 2011

In designing multi-core processors, interconnection delay is one of the major constraints in performance improvement. To solve this problem, the 3-dimensional integration technology has been adopted in designing multi-core processors. The 3D multi-core architecture can reduce the physical wire length by stacking cores vertically, leading to reduced interconnection delay and reduced power consumption. However, the power density of 3D multi-core architecture is increased significantly compared to the traditional 2D multi-core architecture, resulting in the increased temperature of the processor. In this paper, the floorplan methods which change the forms of vertical placement of the core and the level-2 cache are analyzed to solve the thermal problems in 3D multi-core processors. According to the experimental results, it is an effective way to reduce the temperature in the processor that the core and the level-2 cache are stacked adjacently. Compared to the floorplan where cores are stacked adjacently to each other, the floorplan where the core is stacked adjacently to the level-2 cache can reduce the temperature by 22% in the case of 4-layers, and by 13% in the case of 2-layers.

• Journal of the Korea Society of Computer and Information
• /
• v.15 no.11
• /
• pp.1-9
• /
• 2010

As the process technology scales down, an interconnection has became a major performance constraint for multi-core processors. Recently, in order to mitigate the performance bottleneck of the interconnection for multi-core processors, a 3D integration technique has drawn quite attention. The 3D integrated multi-core processor has advantage for reducing global wire length, resulting in a performance improvement. However, it causes serious thermal problems due to increased power density. For this reason, to design efficient 3D multi-core processors, thermal-aware design techniques should be considered. In this paper, we analyze the temperature on the 3D multi-core processors in function unit level through various experiments. We also present temperature characteristics by varying application features, cooling characteristics, and frequency levels on 3D multi-core processors. According to our experimental results, following two rules should be obeyed for thermal-aware 3D processor design. First, to optimize the thermal profile of cores, the core with higher cooling efficiency should be clocked at a higher frequency. Second, to lower the temperature of cores, a workload with higher thermal impact should be assigned to the core with higher cooling efficiency.