• Journal of KIISE
• /
• v.41 no.11
• /
• pp.871-877
• /
• 2014

Cache replacement algorithms have been developed in order to reduce miss counts. In modern processors, the performance gap between the processor and main memory has been increasing, creating a more important role for cache replacement policies. The Least Recently Used (LRU) policy is one of the most common policies used in modern processors. However, recent research has shown that the performance gap between the LRU and the theoretical optimal replacement algorithm (OPT) is large. Although LRU replacement has been proven to be adequate over and over again, the OPT/LRU performance gap is continuously widening as the cache associativity becomes large. In this study, we observed that there is a potential chance to improve cache performance based on existing LRU mechanisms. We propose a method that enhances the performance of the LRU replacement algorithm based on the access proportion among the lines in a cache set during a period of two successive replacement actions that make the final replacement action. Our experimental results reveals that the proposed method reduced the average miss rate of the baseline 512KB L2 cache by 15 percent when compared to conventional LRU. In addition, the performance of the processor that applied our proposed cache replacement policy improved by 4.7 percent over LRU, on average.

• Korean Chemical Engineering Research
• /
• v.47 no.6
• /
• pp.715-719
• /
• 2009

The phase change electrode board for the bio-information detection through electrical property response of phase change material was developed in this study. We manufactured the electrode board using Aluminum first that is widely used in conventional semiconductor device process. Without further treatment, these aluminum electrodes tend to contain voids in PETEOS(plasma enhanced tetraethyoxysilane) material that are easily detected by cross-sectional SEM(Scanning Electron Microscope). The voids can be easily attacked and transformed into holes in between PETEOS and electrodes after etch back and washing process. In order to resolve this issue of Al electrode board, we developed a electrode board manufacturing method using low resistivity TiN, which has advantages in terms of the step-coverage of phase change($Ge_2Sb_2Te_5$, GST) thin film as well as thermodynamic stability, without etch back and washing process. This TiN material serves as the top and bottom electrode in PRAM(Phase-change Random Access Memory). The good connection between the TiN electrode and GST thin film was confirmed by observing the cross-section of TiN electrode board using SEM. The resistances of amorphous and crystalline GST thin film on TiN electrodes were also measured, and 1000 times difference between the amorphous and crystalline resistance of GST thin film was obtained, which is well enough for the signal detection.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.38 no.9
• /
• pp.611-615
• /
• 2001

Y$_2$O$_3$ thin films have been proposed as a buffering insulator of metal/ferroelectric/insulator/semiconductor field effect transistor(MFISFET)-type ferroelectric random access memory (FRAM). In this study, $Y_2$O$_3$ thin films were etched with inductively coupled plasma(ICP). The etch rates of $Y_2$O$_3$ and YMnO$_3$, and the selectivity of $Y_2$O$_3$ to YMnO$_3$ were investigated by varying Cl$_2$/(Cl$_2$＋Ar) gas mixing ratio. The maximum etch rate of $Y_2$O$_3$, and the selectivity of $Y_2$O$_3$ to YMnO$_3$ were 302$\AA$/min, and 2.4 at Cl$_2$/(Cl$_2$＋Ar) gas mixing ratio of 0.2 respectively. Optical emission spectroscopy(OES) was used to understand the effects of gas combination on the etch rate of $Y_2$O$_3$ thin film. The surface reaction of the etched $Y_2$O$_3$ thin films was investigated by x-ray photoelectron spectroscopy (XPS). XPS analysis confirmed that there was chemical reaction between Y and Cl. This result was confirmed by secondary ion mass spectroscopy(SIMS) analysis.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.47 no.4
• /
• pp.69-74
• /
• 2010

A new package substrate for dynamic random access memory(DRAM) devices has been developed using selective aluminum anodization. Unlike the conventional substrate structure commonly made by laminating epoxy-based core and copper clad, this substrate consists of bottom aluminum, middle anodic aluminum oxide and top copper. Anodization process on the aluminum substrate provides thick aluminum oxide used as a dielectric layer in the package substrate. Placing copper traces on the anodic aluminum oxide layer, the resulting two-layer metal structure is completed in the package substrate. Selective anodization process makes it possible to construct a fully filled via structure. Also, putting vias directly in the bonding pads and the ball pads in the substrate design, via in pad structure is applied in this work. These arrangement of via in pad and two-layer metal structure make routing easier and thus provide more design flexibility. In a substrate design, all signal lines are routed based on the transmission line scheme of finite-width coplanar waveguide or microstrip with a characteristic impedance of about $50{\Omega}$ for better signal transmission. The property and performance of anodic alumina based package substrate such as layer structure, design method, fabrication process and measurement characteristics are investigated in detail.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2014.02a
• /
• pp.363-363
• /
• 2014

Ruthenium (Ru) has attractive material properties due to its promising characteristics such as a low resistivity ($7.1{\mu}{\Omega}{\cdot}cm$ in the bulk), a high work function of 4.7 eV, and feasibility for the dry etch process. These properties make Ru films appropriate for various applications in the state-of-art semiconductor device technologies. Thus, it has been widely investigated as an electrode for capacitor in the dynamic random access memory (DRAM), a metal gate for metal-oxide semiconductor field effect transistor (MOSFET), and a seed layer for Cu metallization. Due to the continuous shrinkage of microelectronic devices, better deposition processes for Ru thin films are critically required with excellent step coverages in high aspect ratio (AR) structures. In these respects, atomic layer deposition (ALD) is a viable solution for preparing Ru thin films because it enables atomic-scale control of the film thickness with excellent conformality. A recent investigation reported that the nucleation of ALD-Ru film was enhanced considerably by using a zero-valent metallorganic precursor, compared to the utilization of precursors with higher metal valences. In this study, we will present our research results on the synthesis and characterization of novel ruthenium complexes. The ruthenium compounds were easy synthesized by the reaction of ruthenium halide with appropriate organic ligands in protic solvent, and characterized by NMR, elemental analysis and thermogravimetric analysis. The molecular structures of the complexes were studied by single crystal diffraction. ALD of Ru film was demonstrated using the new Ru metallorganic precursor and O2 as the Ru source and reactant, respectively, at the deposition temperatures of $300-350^{\circ}C$. Self-limited reaction behavior was observed as increasing Ru precursor and O2 pulse time, suggesting that newly developed Ru precursor is applicable for ALD process. Detailed discussions on the chemical and structural properties of Ru thin films as well as its growth behavior using new Ru precursor will be also presented.

• The KIPS Transactions:PartD
• /
• v.10D no.1
• /
• pp.1-12
• /
• 2003

Recently, there have been many indexing schemes for multimedia data such as image, video data. But recent database applications, for example data mining and multimedia database, are required to support multi-user environment. In order for indexing schemes to be useful in multi-user environment, a concurrency control algorithm is required to handle it. So we propose a concurrency control algorithm that can be applied to CBF (cell-based filtering method), which uses the signature of the cell for alleviating the dimensional curse problem. In addition, we extend the SHORE storage system of Wisconsin university in order to handle high-dimensional data. This extended SHORE storage system provides conventional storage manager functions, guarantees the integrity of high-dimensional data and is flexible to the large scale of feature vectors for preventing the usage of large main memory. Finally, we implement the web-based image retrieval system by using the extended SHORE storage system. The key feature of this system is platform-independent access to the high-dimensional data as well as functionality of efficient content-based queries. Lastly. We evaluate an average response time of point query, range query and k-nearest query in terms of the number of threads.

• Journal of the Institute of Convergence Signal Processing
• /
• v.15 no.2
• /
• pp.70-75
• /
• 2014

Recently, the demand for high-integrated, low-powered, and high-powered SoC design has been increasing due to the multi-functionality and the miniaturization of digital devices and the high capacity of service informations. With the rapid evolution of the system, the required hardware performances have become diversified, the FPGA system has been increasingly adopted for the rapid verification, and SoC system using the FPGA and the ARM core for control has been growingly chosen. While the AXI bus is used in these kinds of systems in various ways, it is traditionally designed with AXI slave structure. In slave structure, there are problems with the CPU resources because CPU is continually involved in the data transfer and can't be used in other jobs, and with the decreased transmission efficiency because the time not used of AXI bus beomes longer. In this paper, an efficient AXI master interface is proposed to solve this problem. The simulation results show that the proposed system achieves reductions in the consumption clock by an average of 51.99% and in the slice by 31% and that the maximum operating frequency is increased to 107.84MHz by about 140%.

• Proceedings of the Materials Research Society of Korea Conference
• /
• 2012.05a
• /
• pp.82.1-82.1
• /
• 2012

란타늄 산화물 ($La_2O_3$) 박막은 하프늄 산화물 ($HfO_2$) 박막보다 높은 유전 상수와 높은 밴드 오프셋으로 인해 dynamic random access memory(DRAM)에서 유전체 재료로써 연구되어 왔다. 그리고 Lanthanum이 도핑된 HfO2이 더 높은 유전 상수와 낮은 누설 전류 밀도를 갖는 다는 사실이 이전에 보고 된 바 있다. 본 연구에서 우리는 ALD를 이용하여, TiN 하부 전극 위에 $La_2O_3$의 위치를 달리하는 $La_2O_3/HfO_2$의 나노 층상조직 구조(두께 10 nm)를 금속 - 절연체 - 금속 (MIM) 구조로 제작 하였다. ALD는 좋은 comformality와 넓은 지역 균일성을 가지며, 원자수준의 두께를 조절할 수 있다는 장점을 갖고 있다. 또한, 다양한 화학 물질들을 이용한 복합적 계층구조를 만들 수 있는 점과 $HfO_2$ 및 $La_2O_3$ 계층의 수직 위치를 정확하게 조절할 수 있는 점으로 본 연구에 적합한 증착 방법이다. HfO2 속에 $La_2O_3$ 층을 깊이에 따라 삽입함으로써 $HfO_2$ 계층에 La 도핑의 효과와 더불어 TiN 하부 전극 위의 $La_2O_3$과 $HfO_2$의 차이점을 확인 하였다. $HfO_2$은 $250^{\circ}C$에서 TDMAH와 물을 사용하여, $La_2O_3$은 동일한 온도에서 $La(iPrCp)_3$와 물을 사용하여 제작되었다. 화학적 구성 및 binding 구조는 X선 광전자 분광법 (XPS)을 통해 분석하였다. 전기적 특성(유전 상수 및 누설 전류)은 Capacitance-Voltage (CV)와 Current-Voltage (IV) 측정으로 확인하였다. 결과적으로, $La_2O_3$ 또는 $HfO_2$을 한 종류만 사용한 절연층의 전기적 특성보다, $La_2O_3/HfO_2$의 나노 층상조직 구조가 더 나은 특성 (누설 전류 밀도 : $5.5{\times}10^{-7}\;A/cm^2$ @-1MV/cm, EOT : 14.6)을 갖는다는 것을 확인했고, 더불어 $La_2O_3$의 흡습 성질로 인한 화학 구조와 전기적 특성의 일부 차이를 확인하였다. 본 연구에서는 $HfO_2$ 속에 $La_2O_3$층이 TiN 하부 전극 바로 위에 위치할 때, 즉, 공기 중에 노출되지 않은 $La_2O_3/HfO_2$ 구조에서 가장 좋은 특성의 MIM capacitor를 얻을 수 있었다.

• Journal of KIISE:Computing Practices and Letters
• /
• v.9 no.1
• /
• pp.13-24
• /
• 2003

MPEG-4 is a standard for multimedia applications and provides a set of technologies to satisfy the needs of authors, service providers and end users alike. In this paper, we suggest a Real-time Multiple Circular Buffer (M4RM Buffer) model, which is suitable for streaming these MPEG-4 contents efficiently. M4RM buffer generates each structure of the buffer, which matches well with each object composing an MPEG-4 content, according to the transferred information, and manipulates multiple read/write operations only by its reference. It divides the decoder buffer and the composition buffer, which are described in the standard, by the unit of frame allocated to minimize the range of access. This buffer unit of a frame is allocated according to the object description. Also, it processes the objects synchronization within the buffer and provides APIs for an efficient buffer management to process the real-time user events. Based on the performance evaluation, we show that M4RM buffer model decreases the waiting time in a buffer frame, and so allows the real-time streaming of an MPEG-4 content using the smaller size of the memory block than IM1-2D and Window Media Player.

• KIPS Transactions on Computer and Communication Systems
• /
• v.6 no.3
• /
• pp.105-112
• /
• 2017

PCI Express is a bus technology that connects the processor and the peripheral I/O devices that widely used as an industry standard because it has the characteristics of high-speed, low power. In addition, PCI Express is system interconnect technology such as Ethernet and Infiniband used in high-performance computing and computer cluster. PGAS(partitioned global address space) programming model is often used to implement the one-sided RDMA(remote direct memory access) from multi-host systems, such as computer clusters. In this paper, we design and implement a OpenSHMEM API based on PCI Express maintaining the existing features of OpenSHMEM to implement RDMA based on PCI Express. We perform experiment with implemented OpenSHMEM API through a matrix multiplication example from system which PCs connected with NTB(non-transparent bridge) technology of PCI Express. The PCI Express interconnection network is currently very expensive and is not yet widely available to the general public. Nevertheless, we actually implemented and evaluated a PCI Express based interconnection network on the RDK evaluation board. In addition, we have implemented the OpenSHMEM software stack, which is of great interest recently.