A 4.5-GHz 130-nm 32-KB L0 cache with a leakage-tolerant self reverse-bias bitline scheme

This paper describes a 32-KB two-read, one-write ported L0 cache for 4.5-GHz operation in 1.2-V 130-nm dual-V/sub TH/ CMOS technology. The local bitline uses a leakage-tolerant self reverse-bias (SRB) scheme with nMOS source-follower pullup access transistors, while preserving robust full-swing operation. Gate-source underdrive of -220 mV on the bitline read-select transistors is established without external bias voltages or gate-oxide overstress. Device-level measurements in the 130-nm technology show 72/spl times/ bitline active leakage reduction, enabling low-V/sub TH/ usage, 40% bitline keeper downsizing, and 16 bitcells/bitline. 11% faster read delay and 2/spl times/ higher dc noise robustness are achieved compared with high-performance dual-V/sub TH/ bitline scheme. Sustained performance and robustness benefits of the SRB technique against conventional dynamic bitline with scaling to 100- and 70-nm technology is also presented.     

Ultrasonic nondestructive evaluation of fibre-reinforced composite materials — a review

A framework is presented for examining the effectiveness of limiting and modelling arguments used in the analysis of fluid flows. It consists of examining the consequences of the arguments on the flow problem as a whole and breaking down the limiting/modelling process into a sequence of steps and associated sequence of flow problems, termed here as ‘lidels’. The notion of validity of lidels is given. Several examples are given to explain the present approach. Invited general lecture at 3rd ACFM, Tokyo, September 1986     

Dislocations and flux pinning in YBa2Cu3O7

The dislocation structures of bulk textured and epitaxial thin film YBa₂Cu₃O₇ superconductors are examined. Correlations between increases in flux pinning and dislocation densities are noted. A model for flux pinning by individual dislocations is presented. This gives a treatment of strain induced effects and effects of normal state region interactions. It is shown that the values of pinning predicted are in line with experimental observations.     

Liquid phase epitaxial growth of pure and doped GaSb layers: morphological evolution and native defects

Undoped and Te-doped gallium antimonide (GaSb) layers have been grown on GaSb bulk substrates by the liquid phase epitaxial technique from Ga-rich and Sb-rich melts. The nucleation morphology of the grown layers has been studied as a function of growth temperature and substrate orientation. MOS structures have been fabricated on the epilayers to evaluate the native defect content in the grown layers from theC-V characteristics. Layers grown from antimony rich melts always exhibitp-type conductivity. In contrast, a type conversion fromp- ton- was observed in layers grown from gallium rich melts below 400 C. The electron mobility of undopedn-type layers grown from Ga-rich melts and tellurium doped layers grown from Sb- and Ga-rich solutions has been evaluated. Paper presented at the poster session of MRSI AGM VI. Kharagpur. 1995     

Theoretical Investigation of Metal Cladding for Nanowire and Cylindrical Micropost Lasers

We investigate the transverse modal properties of cylindrical subwavelength metal-clad nanowire and micropost lasers via rigorous theoretical waveguide analysis, including the effects of finite thickness metal cladding and gain in the core. The results of this analysis show that air-metal surface guided TM₀₁ and some hybrid surface guided modes suffer less loss but are less confined to the core, while core-metal surface guided modes are better confined to the core but suffer greater loss. An increase in the thickness of the metal cladding reduces the loss of the core-metal surface guided modes. The modal gain and confinement of the metal-clad cavity are compared to an unclad cavity.     

Unconstrained and constrained motion control of a planar two-link structurally flexible robotic manipulator

Unconstrained and constrained motion control of a planar two-link structurally-flexible robotic manipulator are considered in this study. The dynamic model is obtained by using the extended Hamilton's principle and the Galerkin criterion. A method is presented to obtain the linearized equations of motion in Cartesian space for use in designing the control system. The approach to solving the control problem is to use feedforward and feedback control torques. The feedforward torques maneuver the flexible manipulator along a nominal trajectory and the feedback torques minimize any deviations from the nominal trajectory. The feedforward and feedback torques are obtained by solving the inverse dynamics problem for the rigid manipulator and designing linear quadratic Gaussian with loop transfer recovery (LQG/LTR) compensators, respectively. The LQG/LTR design methodology is exploited to design a robust feedback control system that can handle modeling errors and sensor noise, and operate on Cartesian space trajectory errors. Computer simulated results are presented for an example planar, two-link, structurally flexible robotic manipulator. © 1994 John Wiley & Sons, Inc.     

High-utility and diverse itemset mining

High-utility Itemset Mining (HUIM) finds patterns from a transaction database with their utility no less than a user-defined threshold. The utility of an itemset is defined as the sum of the utilities of its items. The utility notion enables a data analyst to associate a profit score with each item and thereof to a pattern. We extend the notion of high-utility with diversity to define a new pattern type called High-utility and Diverse pattern (HUD). The notion of diversity of a pattern captures the extent of the different categories covered by the selected items in the pattern. An application of diverse-pattern lies in the recommendation task where a system can recommend to a customer a set of items from a new class based on her previously bought items. Our notion of diversity is easy to compute and also captures the basic essence of a previously proposed diversity notion. The existing algorithm to compute frequent-diverse patterns is 2-phase, i.e., in the first phase, frequent patterns are computed, out of which diverse patterns are filtered out in the second phase. We, in this paper, give an integrated algorithm that efficiently computes high-utility and diverse patterns in a single phase. Our experimental study shows that our proposed algorithm is very efficient as compared to a 2-phase algorithm that extracts high-utility itemsets in the first phase and filters out the diverse itemsets in the second phase.