Geri Dön

Variability and reliability in nanoscale circuits: Simulation, design, monitoring, and characterization

Başlık çevirisi mevcut değil.

  1. Tez No: 401060
  2. Yazar: MUSTAFA BERKE YELTEN
  3. Danışmanlar: DR. MICHEAL B. STEER, DR. PAUL D. FRANZON
  4. Tez Türü: Doktora
  5. Konular: Elektrik ve Elektronik Mühendisliği, Electrical and Electronics Engineering
  6. Anahtar Kelimeler: Belirtilmemiş.
  7. Yıl: 2011
  8. Dil: İngilizce
  9. Üniversite: North Carolına State Unıversıty
  10. Enstitü: Yurtdışı Enstitü
  11. Ana Bilim Dalı: Belirtilmemiş.
  12. Bilim Dalı: Belirtilmemiş.
  13. Sayfa Sayısı: Belirtilmemiş.

Özet

Özet yok.

Özet (Çeviri)

At the nanoscale, variability of transistor process parameters and time-based reliability degradation (also called device aging) significantly impact circuit performance metrics. It is necessary to individually model the variability of each transistor in a circuit to accurately represent die-to-die and within-die variations. Traditionally, one of several sets of process parameters is used for all transistors in a circuit in what is called process corner analysis. Reduced-order models are developed to rapidly analyze and simulate circuit variability and reliability simultaneously. The main advantage of reduced-order models is integrating multiple parameter sweeps by greatly reducing the dimensionality of a transistor process model to just those parameters subject to significant variability. The surrogate models also include models of aging that are not part of a transistor model. Other effects could also be incorporated. The drain-source currents of 65 nm n- and p- channel transistors are modeled using surrogate models, neural networks, and support vector machines. The models are in terms of eleven parameters, including a few process parameters that essentially cover the process variation space, terminal voltages, temperature and device age. The reducedorder models are compared on the basis of the model generation and evaluation times, model size, and accuracy. Kriging-based surrogate models and neural network-based models are shown to have sufficient accuracy to perform circuit variability and reliability analysis with Kriging-based models having a shorter evaluation time for smaller circuits and neural network-based models having shorter evaluation time for larger circuits. Variability in operating conditions of differential amplifiers is accurately captured using the reduced-order models. The effects of die-to-die and within-die variations on voltage transfer characteristics of an XOR gate are presented. Individual contributions of crucial process parameters to drain current mismatch of two transistors are quantified with the largest contributions coming from intrinsic threshold voltages and effective length variations of devices. Reliability degradation in differential amplifiers in the presence of process variations are investigated. Design for reliability guidelines are formulated suggesting that the biasing of analog circuits should be modulated by changing nodal voltages within a circuit to increase device aging resilience. An analog circuit-based negative bias temperature monitor is designed and implemented in 65 nm device technology yielding a sensitivity of 3.15 V−1. Finally, experimental aging characterization examples have been conducted on major analog and digital building blocks.

Benzer Tezler

  1. The impact of rating methods and rater training on the variability and reliability of EFL students' classroom-based writing assessments in Turkish universities: An investigation of problems and solutions

    Puanlama yöntemlerinin ve puanlayıcı eğitiminin Türkiyede üniversilerdeki EFL öğrencilerinin sınıf temelli yazma değerlendirmelerinin değişkenliği ve güvenirliği üzerine etkisi: Bir problemler ve çözümler araştırması

    TURGAY HAN

    Doktora

    İngilizce

    İngilizce

    2013

    Eğitim ve ÖğretimAtatürk Üniversitesi

    İngiliz Dili ve Edebiyatı Ana Bilim Dalı

    DOÇ. DR. JINYAN HUANG

    YRD. DOÇ. DR. HÜSEYİN EFE

  2. Characterization of time-based degradation effects and machine learning-based modeling of hot carrier injection in 40 NM CMOS transistors

    40 NM CSMOS transistörlerde sıcak taşıyıcı enjeksiyonunun zaman bazlı bozulma etkilerinin karakterizasyonu ve makine öğrenimine dayalı modellenmesi

    XHESİLA XHAFA

    Yüksek Lisans

    İngilizce

    İngilizce

    2021

    Elektrik ve Elektronik Mühendisliğiİstanbul Teknik Üniversitesi

    Elektronik ve Haberleşme Mühendisliği Ana Bilim Dalı

    DOÇ. DR. MUSTAFA BERKE YELTEN

  3. Transactive energy in electrical distribution network market structure and reliability analysis

    Elektrik dağıtım şebekesinde transaktif enerji pazar yapısı ve güvenilirlik analizi

    KEYSAN POLAT

    Doktora

    İngilizce

    İngilizce

    2024

    Elektrik ve Elektronik Mühendisliğiİstanbul Teknik Üniversitesi

    Elektrik Mühendisliği Ana Bilim Dalı

    PROF. DR. AYDOĞAN ÖZDEMİR

  4. Investigating the effectiveness of automated feedback and the impact of automated scoring on the variability and reliability of EFL writing scores

    Otomatik geri bildirimin etkililiğinin ve otomatik puanlamanın İngilizce kompozisyon puanlarının değişkenliğine ve güvenirliğine etkisinin incelenmesi

    ELİF SARI

    Doktora

    İngilizce

    İngilizce

    2020

    Eğitim ve ÖğretimAtatürk Üniversitesi

    Yabancı Diller Eğitimi Ana Bilim Dalı

    PROF. DR. MEHMET TAKKAÇ

    DOÇ. DR. TURGAY HAN

  5. Spinal anestezi altında yapılan periferik varis operasyonlarında, volüm durumunun ultrasonografik vena cava inferior çap ölçümü ve pleth varıabılıty ındex ile değerlendirilmesi

    The evolution of volume status with ultrasound-assisted measurement of diameter of inferior vena cava and pleth variability index in patients undergoing peripheric variceal operations under spinal anesthesia

    İREM DURMUŞ

    Tıpta Uzmanlık

    Türkçe

    Türkçe

    2021

    Anestezi ve ReanimasyonSağlık Bilimleri Üniversitesi

    Anesteziyoloji ve Reanimasyon Ana Bilim Dalı

    UZMAN ŞENAY GÖKSU