Scaling planar silicon devices

The generation-over-generation scaling of critical CMOS technology parameters is ultimately bound by nonscalable limitations, such as the thermal voltage and the elementary electronic charge. Sustained improvement in performance and density has required the introduction of new device structures and materials. Partially depleted SOI, a most recent MOSFET innovation, has extended VLSI performance while introducing unique idiosyncrasies. Fully depleted SOI is one logical extension of this device design direction. Gate dielectric tunneling, device self-heating, and single-event upsets present developers of these next-generation devices with new challenges. Strained silicon and high-permittivity gate dielectric are examples of new materials that will enable CMOS developers to continue to deliver device performance enhancements in the sub-100 nm regime.     

混合信号芯片设计处理关键的转折阶段

通过简单地剪切和粘贴知识产权(IP)内核可以加快无工厂半导体公司的系统级芯片(SOC)设计。 过去十年中,涌现出大量的为系统制造商提供专用芯片(ASIC)的小型IC设计公司。这些被称为无工厂企业(因为他们将IC制造过程转交给商业芯片制造工厂),需要的启动资金较少,而且如果市场接受他们的产品的话,能够获得丰厚的回报。在大量设计工具的支持下,这些无工厂设计企业在历史悠久的大型芯片制造商,如IBM、Intel、Motorola和德州仪器公司所主导的市场中赢得了一席之地。     

A three-phase model for mixed columnar-equiaxed solidification

A three-phase model for mixed columnar-equiaxed solidification is presented in this article. The three phases are the parent melt as the primary phase, as well as the solidifying columnar dendrites and globular equiaxed grains as two different secondary phases. With an Eulerian approach, the three phases are considered as spatially coupled and interpenetrating continua. The conservation equations of mass, momentum, species, and enthalpy are solved for all three phases. An additional conservation equation for the number density of the equiaxed grains is defined and solved. Nucleation of the equiaxed grains, diffusion-controlled growth of both columnar and equiaxed phases, interphase exchanges, and interactions such as mass transfer during solidification, drag force, solute partitioning at the liquid/solid interface, and release of latent heat are taken into account. Binary steel ingots (Fe-0.34 wt pct C) with two-dimensional (2-D) axis symmetrical and three-dimensional (3-D) geometries as a benchmark were simulated. It is demonstrated that the model can be used to simulate the mixed columnar-equiaxed solidification, including melt convection and grain sedimentation, macrosegregation, columnar-to-equiaxed-transition (CET), and macrostructure distribution. The model was evaluated by comparing it to classical analytical models based on limited one-dimensional (1-D) cases. Satisfactory results were obtained. It is also shown that in order to apply this model for industrial castings, further improvements are still necessary concerning some details.     

Dynamics of Interacting OH– and OD– Ions in KCl

OH^– and OD^– defects are known to form rotational tunneling systems in KCl host crystals. We have studied the complex dielectric susceptibility of KCl doped with different concentrations of OH^– and OD^– in a wide range of frequencies and temperatures. Our main result is that there is a transition from coherent single ion tunneling at low defect concentration to an incoherent tunneling motion at high defect concentrations. In addition, we have studied the thermally activated motion of pairwise coupled hydroxide ions in an attempt to obtain information on the microscopic configurations of the different pairs contributing to the dielectric loss. PACS numbers: 61.72.Ji, 77.22-d, 78.30.Ly     

Agility through discipline: a debate

While Beck asserts that agility is only possible through greater discipline on the part of everyone involved, Boehm counters that you don't broaden the definition of "discipline" by rejecting parts of it. Yet, from inside extreme programming, it seems that the only way to achieve the desired results is to view the world in 'both-and" terms instead of "either-or" terms.     

Dynamic thermographic imaging for estimation of regional perfusion in the tuberculin reaction in healthy adults

A sensitive method for measurement of the volume of blood flow through the skin, based on the kinetics of reheating after localised cooling, is described in this paper. This method has been used to study the tuberculin reaction as a model of cutaneous delayed-type hypersensitivity (DHS) in man. Over the positive reaction there is accelerated reheating similar in kinetics and extent to that seen after maximal hyperaemia induced by intradermal injection of histamine or prostaglandin E2. The earlier phase of reheating (10-100 s) is more dependent on blood flow, whereas the later phase (100-300 s) is apparently more dependent on non-perfusion heat exchange mechanisms, including conduction. The reheat kinetic method is largely dependent on blood flow in the deep dermal vessels (diameter > 50 microns), whereas the alternative approach of measurement of the velocity of flow of erythrocytes in the microcirculation by laser Doppler (LD) flowmetry gives results biased towards the most superficial dermal circulation. Previous studies with LD flowmetry have shown that the blood velocity is greatest at the centre of weak and strong reactions, while in the most intense reactions it is raised at the centre but maximal at the periphery (central relative slowing, CRS) raising the possibility of central ischaemia. The reheat kinetics approach has now indicated that the deep dermal circulation is not impaired in CRS reactions. It is concluded that there must be partial obstruction of the parts of the microcirculation communicating between the deep and superficial dermal plexuses, presumably from the accumulation of exudate oedema in the most intense tuberculin reactions.     

CMOS on FZ-high resistivity substrate for monolithic integration ofSiGe-RF-circuitry and readout electronics

The choice of a highly resistive substrate for silicon millimeter-wave integrated circuits (SIMMWIC) imposed by the requirement of low RF-substrate losses requires the adaptation of a CMOS process on float zone silicon (FZ). A comparison of n- and p-channel devices realized on high resistivity substrate (p-type, 5000 Ω·cm) and standard CMOS substrates (CZ, n-type, 4-6 Ω·cm) is given. Using careful process design, we obtained device characteristics on FZ-substrates that are closely similar to those on standard material, thus allowing direct transfer of existing circuit designs