In-situ power monitoring scheme and its application in dynamic voltage and threshold scaling for digital CMOS integrated circuits

An in-situ power monitoring technique for Dynamic Voltage and Threshold scaling (DVTS) systems is proposed which measures total power consumed by load circuit using sleep transistor acting as power sensor. Design details of power monitor are examined using simulation framework in UMC 90nm CMOS process. Experimental results of test chip fabricated in AMS 0.35µm CMOS process are presented. The test chip has variable activity between 0.05 and 0.5 and has PMOS VTH control through nWell contact. Maximum resolution obtained from power monitor is 0.25mV. Overhead of power monitor in terms of its power consumption is 0.244 mW (2.2% of total power of load circuit). Lastly, power monitor is used to demonstrate closed loop DVTS system. DVTS algorithm shows 46.3% power savings using in-situ power monitor.

High-Performance WSe2 CMOS Technology and Integrated Circuits.

Because of their extraordinary structural and electrical properties, two dimensional materials are currently being pursued for applications such as thin-film transistors and integrated circuit. One of the main challenges that still needs to be overcome for these applications is the fabrication of air-stable transistors with industry-compatible complementary metal oxide semiconductor (CMOS) technology. In this work, we experimentally demonstrate a novel high performance air-stable WSe2 CMOS technology with almost ideal voltage transfer characteristic, full logic swing and high noise margin with different supply voltages. More importantly, the inverter shows large voltage gain (~38) and small static power (Pico-Watts), paving the way for low power electronic system in 2D materials.

Impact of energy quantisation in single electron transistor island on hybrid complementary metal oxide semiconductor-single electron transistor integrated circuits

For the first time, the impact of energy quantisation in single electron transistor (SET) island on the performance of hybrid complementary metal oxide semiconductor (CMOS)-SET transistor circuits has been studied. It has been shown through simple analytical models that energy quantisation primarily increases the Coulomb Blockade area and Coulomb Blockade oscillation periodicity of the SET device and thus influences the performance of hybrid CMOS-SET circuits. A novel computer aided design (CAD) framework has been developed for hybrid CMOS-SET co-simulation, which uses Monte Carlo (MC) simulator for SET devices along with conventional SPICE for metal oxide semiconductor devices. Using this co-simulation framework, the effects of energy quantisation have been studied for some hybrid circuits, namely, SETMOS, multiband voltage filter and multiple valued logic circuits. Although energy quantisation immensely deteriorates the performance of the hybrid circuits, it has been shown that the performance degradation because of energy quantisation can be compensated by properly tuning the bias current of the current-biased SET devices within the hybrid CMOS-SET circuits. Although this study is primarily done by exhaustive MC simulation, effort has also been put to develop first-order compact model for SET that includes energy quantisation effects. Finally, it has been demonstrated that one can predict the SET behaviour under energy quantisation with reasonable accuracy by slightly modifying the existing SET compact models that are valid for metallic devices having continuous energy states.

Synchronisation in high-performance integrated circuits

A distribui ção de um sinal relógio, com elevada precisão espacial (baixo skew) e temporal (baixo jitter ), em sistemas sí ncronos de alta velocidade tem-se revelado uma tarefa cada vez mais demorada e complexa devido ao escalonamento da tecnologia. Com a diminuição das dimensões dos dispositivos e a integração crescente de mais funcionalidades nos Circuitos Integrados (CIs), a precisão associada as transições do sinal de relógio tem sido cada vez mais afectada por varia ções de processo, tensão e temperatura. Esta tese aborda o problema da incerteza de rel ogio em CIs de alta velocidade, com o objetivo de determinar os limites do paradigma de desenho sí ncrono. Na prossecu ção deste objectivo principal, esta tese propõe quatro novos modelos de incerteza com âmbitos de aplicação diferentes. O primeiro modelo permite estimar a incerteza introduzida por um inversor est atico CMOS, com base em parâmetros simples e su cientemente gen éricos para que possa ser usado na previsão das limitações temporais de circuitos mais complexos, mesmo na fase inicial do projeto. O segundo modelo, permite estimar a incerteza em repetidores com liga ções RC e assim otimizar o dimensionamento da rede de distribui ção de relógio, com baixo esfor ço computacional. O terceiro modelo permite estimar a acumula ção de incerteza em cascatas de repetidores. Uma vez que este modelo tem em considera ção a correla ção entre fontes de ruí do, e especialmente util para promover t ecnicas de distribui ção de rel ogio e de alimentação que possam minimizar a acumulação de incerteza. O quarto modelo permite estimar a incerteza temporal em sistemas com m ultiplos dom ínios de sincronismo. Este modelo pode ser facilmente incorporado numa ferramenta autom atica para determinar a melhor topologia para uma determinada aplicação ou para avaliar a tolerância do sistema ao ru ído de alimentação. Finalmente, usando os modelos propostos, são discutidas as tendências da precisão de rel ogio. Conclui-se que os limites da precisão do rel ogio são, em ultima an alise, impostos por fontes de varia ção dinâmica que se preveem crescentes na actual l ogica de escalonamento dos dispositivos. Assim sendo, esta tese defende a procura de solu ções em outros ní veis de abstração, que não apenas o ní vel f sico, que possam contribuir para o aumento de desempenho dos CIs e que tenham um menor impacto nos pressupostos do paradigma de desenho sí ncrono.

Silicon Photonics Integrated Circuits for Flexible Optical Systems

This dissertation deals with the design and the characterization of novel reconfigurable silicon-on-insulator (SOI) devices to filter and route optical signals on-chip. Design is carried out through circuit simulations based on basic circuit elements (Building Blocks, BBs) in order to prove the feasibility of an approach allowing to move the design of Photonic Integrated Circuits (PICs) toward the system level. CMOS compatibility and large integration scale make SOI one of the most promising material to realize PICs. The concepts of generic foundry and BB based circuit simulations for the design are emerging as a solution to reduce the costs and increase the circuit complexity. To validate the BB based approach, the development of some of the most important BBs is performed first. A novel tunable coupler is also presented and it is demonstrated to be a valuable alternative to the known solutions. Two novel multi-element PICs are then analysed: a narrow linewidth single mode resonator and a passband filter with widely tunable bandwidth. Extensive circuit simulations are carried out to determine their performance, taking into account fabrication tolerances. The first PIC is based on two Grating Assisted Couplers in a ring resonator (RR) configuration. It is shown that a trade-off between performance, resonance bandwidth and device footprint has to be performed. The device could be employed to realize reconfigurable add-drop de/multiplexers. Sensitivity with respect to fabrication tolerances and spurious effects is however observed. The second PIC is based on an unbalanced Mach-Zehnder interferometer loaded with two RRs. Overall good performance and robustness to fabrication tolerances and nonlinear effects have confirmed its applicability for the realization of flexible optical systems. Simulated and measured devices behaviour is shown to be in agreement thus demonstrating the viability of a BB based approach to the design of complex PICs.

Al/SiC carriers for microwave integrated circuits by a new technique of pressureless infiltration

Materials with high thermal conductivity and thermal expansion coefficient matching with that of Si or GaAs are being used for packaging high density microcircuits due to their ability of faster heat dissipation. Al/SiC is gaining wide acceptance as electronic packaging material due to the fact that its thermal expansion coefficient can be tailored to match with that of Si or GaAs by varying the Al:SiC ratio while maintaining the thermal conductivity more or less the same. In the present work, Al/SiC microwave integrated circuit (MIC) carriers have been fabricated by pressureless infiltration of Al-alloy into porous SiC preforms in air. This new technique provides a cheaper alternative to pressure infiltration or pressureless infiltration in nitrogen in producing Al/SiC composites for electronic packaging applications. Al-alloy/65vol% SiC composite exhibited a coefficient of thermal expansion of 7 x 10(-6) K-1 (25 degrees C-100 degrees C) and a thermal conductivity of 147 Wm(-1) K-1 at 30 degrees C. The hysteresis observed in thermal expansion coefficient of the composite in the temperature range 100 degrees C-400 degrees C has been attributed to the presence of thermal residual stresses in the composite. Thermal diffusivity of the composite measured over the temperature range from 30 degrees C to 400 degrees C showed a 55% decrease in thermal diffusivity with temperature. Such a large decrease in thermal diffusivity with temperature could be due to the presence of micropores, microcracks, and decohesion of the Al/SiC interfaces in the microstructure (all formed during cooling from the processing temperature). The carrier showed satisfactory performance after integrating it into a MIC.

Dynamic Supply and Threshold Voltage Scaling for CMOS Digital Circuits Using In-Situ Power Monitor

A generalized power tracking algorithm that minimizes power consumption of digital circuits by dynamic control of supply voltage and the body bias is proposed. A direct power monitoring scheme is proposed that does not need any replica and hence can sense total power consumed by load circuit across process, voltage, and temperature corners. Design details and performance of power monitor and tracking algorithm are examined by a simulation framework developed using UMC 90-nm CMOS triple well process. The proposed algorithm with direct power monitor achieves a power savings of 42.2% for activity of 0.02 and 22.4% for activity of 0.04. Experimental results from test chip fabricated in AMS 350 nm process shows power savings of 46.3% and 65% for load circuit operating in super threshold and near sub-threshold region, respectively. Measured resolution of power monitor is around 0.25 mV and it has a power overhead of 2.2% of die power. Issues with loop convergence and design tradeoff for power monitor are also discussed in this paper.

Dynamic Supply and Threshold Voltage Scaling for CMOS Digital Circuits Using In-Situ Power Monitor

A generalized power tracking algorithm that minimizes power consumption of digital circuits by dynamic control of supply voltage and the body bias is proposed. A direct power monitoring scheme is proposed that does not need any replica and hence can sense total power consumed by load circuit across process, voltage, and temperature corners. Design details and performance of power monitor and tracking algorithm are examined by a simulation framework developed using UMC 90-nm CMOS triple well process. The proposed algorithm with direct power monitor achieves a power savings of 42.2% for activity of 0.02 and 22.4% for activity of 0.04. Experimental results from test chip fabricated in AMS 350 nm process shows power savings of 46.3% and 65% for load circuit operating in super threshold and near sub-threshold region, respectively. Measured resolution of power monitor is around 0.25 mV and it has a power overhead of 2.2% of die power. Issues with loop convergence and design tradeoff for power monitor are also discussed in this paper.

Self Repair in Circuits-Automating Open Fault Repair in Integrated Circuits Using Field-Induced Aggregation of Carbon Nanotubes

This paper presents studies on the use of carbon nanotubes dispersed in an insulating fluid to serve as an automaton for healing open-circuit interconnect faults in integrated circuits. The physics behind the repair mechanism is the electric-field-induced diffusion limited aggregation. On the occurrence of an open fault, the repair is automatically triggered due to the presence of an electric field across the gap. We perform studies on the repair time as a function of the electric field and dispersion concentrations with the above application in mind.