Nanometer CMOS ICs From Basics to

ASICs 2nd Edition Harry J.M. Veendrick

(Auth.)
Visit to download the full and correct content document:
https://textbookfull.com/product/nanometer-cmos-ics-from-basics-to-asics-2nd-edition
-harry-j-m-veendrick-auth/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Bits on Chips Harry Veendrick

https://textbookfull.com/product/bits-on-chips-harry-veendrick/

Deluge From Genesis to Atlantis 2019 docx Harry

Sivertsen

https://textbookfull.com/product/deluge-from-genesis-to-
atlantis-2019-docx-harry-sivertsen/

Design of Analog CMOS Integrated Circuits 2nd eddition

2nd Edition Behzad Razavi

https://textbookfull.com/product/design-of-analog-cmos-
integrated-circuits-2nd-eddition-2nd-edition-behzad-razavi/

Flash Lamp Annealing From Basics to Applications Lars

Rebohle

https://textbookfull.com/product/flash-lamp-annealing-from-
basics-to-applications-lars-rebohle/
Design of CMOS phase locked loops from circuit level to
architecture level 1st Edition Behzad Razavi

https://textbookfull.com/product/design-of-cmos-phase-locked-
loops-from-circuit-level-to-architecture-level-1st-edition-
behzad-razavi/

Simpler Living Back to Basics Guides 2nd Edition Jeff

Davidson

https://textbookfull.com/product/simpler-living-back-to-basics-
guides-2nd-edition-jeff-davidson/

COVID 19 From Basics To Clinical Practice 1st Edition

Wenhong Zhang

https://textbookfull.com/product/covid-19-from-basics-to-
clinical-practice-1st-edition-wenhong-zhang/

lacanian ink 29 From an Other to the other Josefina

Ayerza

https://textbookfull.com/product/lacanian-ink-29-from-an-other-
to-the-other-josefina-ayerza/

Deep Learning Vol 2 From Basics to Practice Andrew

Glassner

https://textbookfull.com/product/deep-learning-vol-2-from-basics-
to-practice-andrew-glassner/
Harry J.M. Veendrick

Nanometer
CMOS ICs
From Basics to ASICs
Second Edition
Nanometer CMOS ICs
Harry J.M. Veendrick

Nanometer CMOS ICs

From Basics to ASICs

Second Edition

123
Harry J.M. Veendrick
Heeze, The Netherlands

ISBN 978-3-319-47595-0 ISBN 978-3-319-47597-4 (eBook)

DOI 10.1007/978-3-319-47597-4

Library of Congress Control Number: 2016963634

© Springer Netherlands My Business Media 2008

© Springer International Publishing AG 2017
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

Illustrations created by Kim Veendrick and Henny Alblas

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Foreword

CMOS scaling has entered the sub-20 nm era. This enables the design of system-on-
a-chip containing more than ten billion transistors. However, nanometre level device
physics also causes a plethora of new challenges that percolate all the way up to the
system level. Therefore, system-on-a-chip design is essentially teamwork requiring
a close dialogue between system designers, software engineers, chip architects,
intellectual property providers, and process and device engineers. This is hardly
possible without a common understanding of the nanometre CMOS medium, its
terminology, its future opportunities and possible pitfalls. This is what this book
provides.
It is a greatly extended and revised version of the previous edition. So besides
the excellent coverage of all basic aspects of MOS devices, circuits and systems, it
leads the reader into the novel intricacies resulting from scaling CMOS towards the
sub-10 nm level. This new edition contains updates and additional information on
the issues of increased leakage power and its mitigation, to strain induced mobility
enhancement. Immersion and double patterning litho and extreme UV and other
alternative litho approaches for sub-20 nm are extensively discussed together with
their impact on circuit layout. The design section now also extensively covers
design techniques for improved robustness, yield and manufacturing in view of
increased device variability, soft errors and decreased reliability when reaching
atomic dimensions. Both devices and ICs have entered the 3D era. This is reflected
by discussions on FinFETs, gate-all-around transistors, 3D memories and stacked
memory dies and 3D packaging to fully enable system-in-a-package solutions.
Finally, the author shares his thoughts on the challenges of further scaling when
approaching the end of the CMOS roadmap somewhere in the next decade.
This book is unique in that it covers in a very comprehensive way all aspects of
the trajectory from state-of-the-art process technology to the design and packaging
of robust and testable systems in nanometre scale CMOS. It is the reflection
of the author’s own research in this domain but also of more than 35 years of
experience in training the full CMOS chip development chain to more than 4500
semiconductor professionals at Philips, NXP, ASML, Infineon, ST Microelectron-
ics, TSMC, Applied Materials, IMEC, etc. It provides context and perspective to all
semiconductor disciplines.

v
vi Foreword

I strongly recommend this book to all engineers involved in the design, lithogra-
phy, manufacturing and testing of future systems-on-silicon as well as to engineer-
ing undergraduates who want to understand the basics that make electronics systems
work.

Senior Fellow IMEC Hugo De Man

Professor Emeritus K.U. Leuven
Leuven, Belgium
January 2017
Preface

An integrated circuit (IC) is a piece of semiconductor material, on which a

number of electronic components are interconnected. These interconnected ‘chip’
components implement a specific function. The semiconductor material is usually
silicon, but alternatives include gallium arsenide. ICs are essential in most modern
electronic products. The first IC was created by Jack Kilby in 1959. Photographs of
this device and the inventor are shown in Fig. 3. Figure 1 illustrates the subsequent
progress in IC complexity. This figure shows the numbers of components for
advanced ICs and the year in which these ICs were first presented. This doubling in
complexity every 2 years was predicted by Moore (Intel 1964), whose law is still
valid today for the number of logic transistors on a chip. However, due to reaching
the limits of scaling, the complexity doubling of certain memories now happens
at a 3-year cycle. This is shown by the complexity growth line which is slowly
saturating.
Figure 2 shows the relative semiconductor revenue per IC category. CMOS
ICs take about more than 80% of the total semiconductor market. Today’s digital
ICs may contain several hundreds of millions to several billion transistors on
one to several 1 cm2 chip. They can be subdivided into three categories: logic,
microprocessors and memories. About 13% of the CMOS ICs are of an analogue
nature.
Figures 4, 5, 6, 7 and 8 illustrate the evolution in IC technology. Figure 4
shows a discrete BC107 transistor. The digital filter shown in Fig. 5 comprises
a few thousand transistors, while the Digital Audio Broadcasting (DAB) chip in
Fig. 6 contains more than six million transistors. The Intel Haswell-E/EP eight-
core processor of the Xeon family in Fig. 7.30 (Sect. 7.6) contains 2.6 billion
transistors on a 355 mm2 die, fabricated in a 22 nm process with a maximum power
consumption of 140 W thermal design power (TDP) . This is the maximum amount
of heat that the chip’s cooling system can dissipate. Another strong workhorse,
shown in Fig. 7, is the GP100 Pascal chip of Nvidia with 3584 stream processors
containing 15.3 billion transistors, which is fabricated in a 16 nm FinFET process
on a 610 mm2 large die and consumes 300 W. Figure 8 shows a 128 Gb TLC NAND
flash, (50 billion transistors), containing wear levelling algorithms to increase
lifetime.

vii
viii Preface

2T
1 Tbit ?
1T expected capacity
256 G 256 Gbit
64 G 64 Gbit
16 G 16 Gbit
4 Gbit
4G
1 Gbit
1G
number of components per IC

256 M 256 Mbit

64 M 64 Mbit
16 M 16 Mbit
4M 4 Mbit
1M 1 Mbit
256 k 256 kbit
64 k 64 kbit
16 k 16-kbit MOS-DRAM
4k
1-kbit MOS-DRAM
1k
256 4-bit TTL-counter
64
dual flip-flop
16
RTL gate
4
SSI MSI LSI VLSI
1
1959 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025
year
Fig. 1 Growth in the number of components per IC

Opto, sensors Bipolar, Analog

& discretes compound MOS ≈ 14 %
≈ 19 % ≈2%

Total Logic
Semicon- MOS ≈ 36 % of
Integrated (Including
ductor MOS Digital
Circuits BiCMOS)
Market ≈ 81 % of Digital Micros
≈ 98 % of the
the Total ≈ 86 % of ≈ 34 % of
integraded
Market MOS MOS Digital
Circuit
Market Memories
≈ 30 % of
MOS Digital

Fig. 2 Relative semiconductor revenue by IC category (Source: IC Insights, 2016)

Preface ix

Fig. 3 The development of the first IC: in 1958, Jack Kilby demonstrated the feasibility of
resistors and capacitors, in addition to transistors, based on semiconductor technology. Kilby,
an employee of Texas Instruments, submitted the patent request entitled ‘Miniaturized Electronic
Circuits’ in 1959. His request was honoured. Recognition by a number of Japanese companies
in 1990 means that Texas Instruments is still benefiting from Kilby’s patent (Source: Texas
Instruments/Koning and Hartman)

Figure 9 illustrates the sizes of various semiconductor components, such as a

silicon atom, a single transistor and an integrated circuit, in perspective. The sizes
of an individual MOS transistor are already similar to those of a virus.
This book provides an insight into all aspects associated with CMOS ICs. The
topics presented include relevant fundamental physics. Technology, design and
implementation aspects are also explained, and applications are discussed. CAD
tools used for the realisation of ICs are described, while current and expected
developments also receive attention.
The contents of this book are based on the CMOS section of an industry-
oriented course entitled ‘An introduction to IC techniques’. The course has been
given almost three decades, formerly in Philips, currently in NXP Semiconductors.
Continuous revision and expansion of the course material ensures that this book is
highly relevant to the IC industry. The level of the discussions makes this book a
suitable introduction for designers, technologists, CAD developers, test engineers,
failure analysis engineers, reliability engineers, technical-commercial personnel and
IC applicants. The text is also suitable for both graduates and undergraduates in
related engineering courses.
Considerable effort has been made to enhance the readability of this book,
and only essential formulae are included. The large number of diagrams and
photographs should reinforce the explanations. The design and application examples
x Preface

Fig. 4 A single BC107

bipolar transistor (Source:
NXP Semiconductors)

are mainly digital. This reflects the fact that more than 85% of all modern CMOS ICs
are digital circuits. However, the material presented will also provide the analogue
designer with a basic understanding of the physics, manufacture and operation of
nanometre CMOS circuits. The chapters are summarised below. For educational
purposes, the first four chapters each start with a discussion on nMOS physics,
nMOS transistor operation, nMOS circuit behaviour, nMOS manufacturing process,
etc. Because the pMOS transistor operation is fully complementary to that of the
nMOS transistor, it is then easier to understand the operation and fabrication of
complementary MOS (CMOS) circuits. The subjects per chapter are chosen in
a very organised and logical sequence so as to gradually build the knowledge,
from basics to ASICs. The knowledge gathered from each chapter is required to
understand the information presented in the next chapter(s). Each chapter ends with
a reference list and exercises. The exercises summarise the important topics of the
chapter and form an important part of the complete learning process.
Chapter 1 contains detailed discussions of the basic principles and fundamental
physics of the MOS transistor. The derivation of simple current-voltage equations
for MOS devices and the explanation of their characteristics illustrate the relation-
ship between process parameters and circuit performance.
Preface xi

Fig. 5 A digital filter which comprises a few thousand transistors (Source: NXP Semiconductors)
xii Preface

Fig. 6 A Digital Audio

Broadcasting (DAB) chip,
which comprises more than
six million transistors
(Source: NXP
Semiconductors)

The continuous reduction of transistor dimensions leads to increased deviation

between the performance predicted by the simple MOS formulae and actual
transistor behaviour. The effects of temperature and the impact of the continuous
scaling of the geometry on this behaviour are explained in Chap. 2. In addition to
their influence on transistor and circuit performance, these effects can also reduce
device lifetime and reliability.
The various technologies for the manufacture of CMOS ICs are examined in
Chap. 3. After a summary on the available different substrates (wafers) used as
starting material, an explanation of the most important associated photolithographic
and processing steps is provided. This precedes a discussion of an advanced
nanometre CMOS technology for the manufacture of modern VLSI circuits.
The design of CMOS circuits is treated in Chap. 4. An introduction to the
performance aspects of nMOS circuits provides an extremely useful background
for the explanation of the CMOS design and layout procedures.
MOS technologies and their derivatives are used to realise the special devices
discussed in Chap. 5. Charge-coupled devices (CCDs), CMOS imagers and MOS
power transistors are among the special devices. Chapter 5 concludes the presenta-
tion of the fundamental concepts behind BICMOS circuit operation.
Stand-alone memories currently represent about 30% of the total semiconductor
market revenue. However, also in logic and microprocessor ICs, embedded mem-
ories represent close to 80% of the total transistor count. So, of all transistors
produced in the world, today, more than 99.5% end up in either a stand-alone or
in an embedded memory. This share is expected to stay at this level or to increase.
Fig. 7 The GP100 Pascal chip of Nvidia with 3,584 stream processors containing 15.3 billion
transistors, fabricated in a 16 nm FinFET process on a 610 mm2 large die, consuming 300 W
and targeted at science-targeted accelerator cards for artificial intelligence and deep-learning
applications, such as used in autonomous cars, automatic image recognition and smart real-time
language translation in video chat applications, for example (Courtesy of Nvidia)
xiv Preface

Fig. 8 A 128 Gb TLC NAND flash (50 billion transistors), containing wear levelling algorithms
to increase lifetime (Courtesy of Micron Technology)

The majority of available memory types are therefore examined in Chap. 6. The
basic structures and the operating principles of various types are explained. In
addition, the relationships between their respective properties and application areas
are made clear.
Developments in IC technology now facilitate the integration of complete
system-on-a-chip, which contain several hundreds of millions to several billion
transistors. The various IC designs and realisation techniques used for these VLSI
ICs are presented in Chap. 7. The advantages and disadvantages of the techniques
and the associated CAD tools are examined. Various modern technologies are used
to realise a separate class of VLSI ICs, which are specified by applicants rather than
manufacturers. These application-specific ICs (ASICs) are examined in this chapter
as well. Motives for their use are also discussed.
As a result of the continuous increase of power consumption, the maximum level
that can be sustained by cheap plastic packages has been reached. Therefore, all
CMOS designers must have a ‘less-power attitude’. Chapter 8 presents a complete
overview of less-power and less-leakage options for CMOS technologies, as well as
for the different levels of design hierarchy.
Increased VLSI design complexities, combined with higher frequencies, create
a higher sensitivity to physical effects. These effects dominate the reliability and
signal integrity of nanometre CMOS ICs. Chapter 9 discusses these effects and the
design measures to be taken to maintain both reliability and signal integrity at a
sufficiently high level.
Preface xv

Fig. 9 Various semiconductor component sizes (e.g. atom, transistor, integrated circuit) in
perspective

Finally, testing, yield, packaging, debug and failure analysis are important factors
that contribute to the ultimate costs of an IC. Chapter 10 presents an overview of the
state-of-the-art techniques that support testing, debugging and failure analysis. It
also includes a rather detailed summary on available packaging technologies and
gives an insight into their future trends. Essential factors related to IC production
are also examined; these factors include quality and reliability.
The continuous reduction of transistor dimensions associated with successive
process generations is the subject of the final chapter (Chap. 11). This scaling has
various consequences for transistor behaviour and IC performance. The resulting
increase of physical effects and the associated effects on reliability and signal
integrity are important topics of attention. The expected consequences of and road-
blocks for further miniaturisation are described. This provides an insight into the
challenges facing the IC industry in the race towards nanometre devices.
xvi Preface

Not all data in this book completely sprouted from my mind. A lot of books
and papers contributed to make the presented material state of the art. Considerable
effort has been made to make the reference list complete and correct. I apologise for
possible imperfections.

Acknowledgements

I wish to express my gratitude to all those who contributed to the realisation

of this book; it is impossible to include all their names. I greatly value my
professional environment: Philips Research Labs, of which the Semiconductor
Research Department is now part of NXP Semiconductors. It offered me the
opportunity to work with many internationally highly valued colleagues who are
all real specialists in their field of semiconductor expertise. Their contributions
included fruitful discussions, relevant texts and manuscript reviews. I would like to
make an exception, here, for my colleagues Marcel Pelgrom and Maarten Vertregt,
who greatly contributed to the discussions held on trends in MOS transistor currents
and variability matters throughout this book, and Roger Cuppens, Maurits Storms
and Roelof Salters for the discussions on non-volatile and random-access memories,
respectively.
I would especially like to thank Andries Scholten and Ronald van Langevelde for
reviewing Chap. 2 and for the discussions on leakage mechanisms in this chapter
and Casper Juffermans and Johannes van Wingerden (both NXP) and Ewoud
Vreugdenhil and Hoite Tolsma (both ASML) for their inputs to and review of
the lithography section in Chap. 3. I would also like to sincerely thank Robert
Lander for his detailed review of the section on CMOS process technologies
and future trends in CMOS devices and Gerben Doornbos for the correct sizes
and doping levels used in the manufacture of state-of-the-art CMOS devices. I
appreciate the many circuit simulations that Octavio Santana has done to create
the tapering factor table in Chap. 4. I am grateful for the review of Chap. 5 on
special circuits and devices based on MOS transistor operation: Albert Theuwissen
(Harvest Imaging) for the section on CCD and image sensors, Johan Donkers and
Erwin Hijzen for the BICMOS section and Jan Sonsky for the high-voltage section.
I also appreciate their willingness to supply me with great photographic material.
Toby Doorn and Ewoud Vreugdenhil are thanked for their review of the memory
chapter (Chap. 6). I appreciate Paul Wielage’s work on statistical simulations with
respect to memory yield loss. I thank Ad Peeters for information on and reviewing
the part on asynchronous design in the low-power chapter (Chap. 8). Reliability is
an important part of Chap. 9, which discusses the robustness of ICs. In this respect,
I want to thank Andrea Scarpa for reviewing the hot-carrier and NBTI subjects,
Frans List and Theo Smedes for the ESD and latch-up subjects and Yuang Li for
the part on electromigration. I also greatly value the work of Bram Kruseman, Henk
Thoonen and Frank Zachariasse for reviewing the sections on testing, packaging
and failure analysis, respectively. I also like to express to them my appreciation for
supplying me with a lot of figures and photographs, which support and enrich the
Preface xvii

discussions on these subjects in Chap. 10. Finally, I want to thank Chris Wyland and
John Janssen, for their remarks and additions on electrical and thermal aspects of IC
packages, respectively
I am very grateful to all those who attended the course, because their feedback
on educational aspects and their corrections and constructive criticism contributed
to the quality and completeness of this book.
In addition, I want to thank Philips Research and NXP Semiconductors, in
general, for the co-operation I was afforded. I thank my son Bram for the layout
of the cover and the layout diagrams in Chap. 4, and Ron Salfrais for the correctness
of a large part of the English text.
I would especially like to express my gratitude to my daughter Kim and Henny
Alblas for the many hours they have spent on the creation of excellent and colourful
art work, which contributes a lot to the quality and clarity of this book.
Finally, I wish to thank Harold Benten and Dré van den Elshout for their
conscientious editing and typesetting work. Their efforts to ensure high quality
should not go unnoticed by the reader.
However, the most important appreciation and gratitude must go to my family,
again, and in particular to my wife, for her years of exceptional tolerance, patience
and understanding. The year 2007 was particularly demanding. Lost hours can never
be regained, but I hope that I can give her now a lot more free time in return.

Eindhoven, The Netherlands Harry J.M. Veendrick

February 2008

This second full-colour edition covers the same subjects, but then they are
completely revised and updated with the most recent material. It covers all subjects,
related to nanometre CMOS ICs: physics, technologies, design, testing, packaging
and failure analysis. The contents include substantially new material along with
extended discussions on existing topics, which leads to a more detailed and complete
description of all semiconductor disciplines. The result is a better self-contained
book which makes it perfectly accessible to semiconductor professionals, academic
staff and PhD and (under)graduate students.
Finally, I wish to thank Harold Benten and Kim Veendrick, again, for their
conscientious text-editing and excellent art work, respectively.

Heeze, The Netherlands Harry J.M. Veendrick

January 2017
Overview of Symbols

˛ Channel-shortening factor or clustering factor

A Area
A Aspect ratio
a Activity factor
ˇ MOS transistor gain factor
ˇ Gain factor for MOS transistor with square channel
ˇn nMOS transistor gain factor
ˇp pMOS transistor gain factor
ˇtotal Equivalent gain factor for a combination of transistors
BV Breakdown voltage
C Capacitance
Cb Bitline capacitance
Cd Depletion layer capacitance
Cdb Drain-substrate capacitance
Cg Gate capacitance
Cgb Gate-substrate capacitance
Cgd Gate-drain capacitance
Cgs Gate-source capacitance
Cgdo Voltage-independent gate-drain capacitance
Cgso Voltage-independent gate-source capacitance
Cpar Parasitic capacitance
Cmin Minimum capacitance
Cs Scaled capacitance
Cox Oxide capacitance
Cs Silicon surface-interior capacitance
Csb Source-substrate (source-bulk) voltage
Ct Total capacitance
CD Critical dimension
L Difference between drawn and effective channel length
VT Threshold-voltage variation
D0 Defect density for uniformly distributed errors (dust particles)
Dl Threshold-voltage channel-length dependence factor
Dw Threshold-voltage channel-width dependence factor

xix
xx Overview of Symbols

 Dielectric constant
0 Absolute permittivity
ox Relative permittivity of oxide
r Relative permittivity
si Relative permittivity of silicon
E Electric field strength
Ec Conduction band energy level
Ef Fermi energy level
Ei Intrinsic (Fermi) energy level
Emx Maximum horizontal electric field strength
Eox Electric field across an oxide layer
Ev Valence band energy level
Ex Horizontal electric field strength
Exc Critical horizontal field strength
Ez Vertical electric field strength
 Electric potential
f Fermi potential
s Surface potential of silicon w.r.t. the substrate interior
MS Contact potential between gate and substrate
F Feature size (= size of a half pitch used for stand-alone memories)
f Clock frequency
fmax Maximum clock frequency
 Factor which expresses relationship between drain-source
voltage and threshold-voltage variation
gm Transconductance
I Current
Ib Substrate current
Ids Drain-source current
Ids0 Characteristic subthreshold current for gate-substrate voltage of 0 V
IdsD Driver transistor drain-source current
IdsL Load transistor drain-source current
Idssat Saturated transistor drain-source current
Idssub Subthreshold drain-source current
Imax Maximum current
Ion On current
IR Current through resistance
i.t/ Time-dependent current
j Current densisty
k Boltzmann’s constant
K K-factor; expresses relationship between
source-substrate voltage and threshold voltage
K Amplification factor
 Wavelength of light
L Effective transistor channel length and inductance
Overview of Symbols xxi

LCLM Channel-length reduction due to channel-length modulation

Leff Effective channel length
Lref Effective channel length of reference transistor
M Yield model parameter
0 Substrate carrier mobility
n Channel electron mobility
p Channel hole mobility
n Number of electrons in a material
ni Ntrinsic carrier concentration
NA Substrate doping concentration
N.A. Numeric aperture
 Charge density
P Power dissipation
Pdyn Dynamic power dissipation
Pstat Static power dissipation
p Voltage scaling factor
p Also represents the number of holes in a material, in related expressions
Q Charge
q Elementary charge of a single electron
Qd Depletion layer charge
Qg Gate charge
Qm Total mobile charge in the inversion layer
Qn Mobile charge per unit area in the channel
Qox Oxide charge
Qs Total charge in the semiconductor
R Resistance
RJA Junction-to-air thermal resistance
RJC Junction-to-case thermal resistance
RL Load resistance
Rout Output resistance or channel resistance
Rtherm Thermal resistance of a package
r Tapering factor
s Scale factor
ssubthr Subthreshold slope
Conductivity of a semiconductor material
Delay time
f Fall time
r Rise time
R Dielectric relaxation time
T Clock period
Tmin Minimum clock period
Temp Temperature
TempA Ambient temperature
TempC Case temperature
xxii Overview of Symbols

TempJ Junction temperature

Tlf Transistor lifetime
t Time
tcond Conductor thickness
td Depletion layer thickness
tdielectric Dielectric thickness
tox Gate-oxide thickness
tis Isolator thickness
U Computing power
v Carrier velocity
vsat Carrier saturation velocity
V Voltage
VB Breakdown voltage
Vr Scaled voltage
V0 Depletion layer voltage
Vbb Substrate voltage
Vdd Supply voltage
Vc Voltage at silicon surface
Vds Drain-source voltage
Vdssat Drain-source voltage of saturated transistor
VE Early voltage
Vfb Flat-band voltage
Vg Gate voltage
Vgg Extra supply voltage
Vgs Gate-source voltage
VgsL Load transistor gate-source voltage
VH High-voltage level
Vin Input voltage
Vj Junction voltage
VL Low-voltage level
VPT Transistor punch-through voltage
Vsb Source-substrate (back-bias) voltage
Vss Ground voltage
Vws Well-source voltage
VT Threshold voltage
VT D Driver transistor threshold voltage
VTdep Depletion transistor threshold voltage
VTenh Enhancement transistor threshold voltage
VT L Load transistor threshold voltage
VT n nMOS transistor threshold voltage
VT p pMOS transistor threshold voltage
VTpar Parasitic transistor threshold voltage
Vout Output voltage
V.x/ Potential at position x
Overview of Symbols xxiii

Vx Process-dependent threshold-voltage term

V XL Process-dependent threshold-voltage term for load transistor
V XD Process-dependent threshold-voltage term for driver transistor
W Transistor channel width
Wn nMOS transistor channel width
Wp pMOS transistor channel width
Wref Reference transistor channel width
W
LW  Transistor aspect ratio
 WL n nMOS transistor aspect ratio
L p
pMOS transistor aspect ratio
x Distance w.r.t. specific reference point
Y Yield
Zi Input impedance
Explanation of Atomic-Scale Terms

Electron: an elementary particle, meaning that it is not built from substructures.

They can be fixed bound to the nucleus of an atom, or freely moving around.
When free electrons move through vacuum or a conductor, they create a flow of
charge. This is called electrical current, which, by definition, flow in the opposite
direction of the negatively charged electrons. Electrons have a mass of 9:11  1031
kg ( 1/1836 the mass of a proton) and a negative charge 1:6  1019 C. Electrons
play a primary role in electronic, magnetic, electromagnetic, chemistry and nuclear
physics. In semiconductor circuits, their main role is charging or discharging analog,
logic and memory nodes.

Proton: a subatomic particle with a positive charge of 1:6  1019 C and a mass
of 1:67  1027 kg. Protons form, together with neutrons, the basic elements from
which all atomic nuclei are built and are held together by a strong nuclear force.

Neutron: a subatomic particle with a no charge, with a mass which is about equal to
that of a proton.

Atom: an atom is the smallest unit of any material in the periodic system of elements.
It consists of a nucleus with a fixed number of protons and neutrons, surrounded by
one or more shells, which each contain a certain number of electrons. Since an atom
is electrically neutral, the total number of electrons in these shells (one or more;
hydrogen has only one electron) is identical to the number of protons in the nucleus,
since neutrons have no net electrical charge. The number of protons in the nucleus
defines the atomic number of the element in the periodic table of elements and
determines their physical and chemical properties and behaviour. Most of the CMOS
circuits are fabricated on silicon wafers. Silicon is in group IV, which means that it
has four electrons in the outer shell. In a mono crystalline silicon substrate, each of
these four electrons can form bonding pairs with corresponding electrons from four
neighbour silicon atoms, meaning that each silicon atom is directly surrounded by
four others. By replacing some of the silicon atoms by boron or phosphorous, one
can change the conductivity of the substrate material In this way nMOS or pMOS
transistors can be created. Atom sizes are of the order of 0:1  0:4 nm. In a mono
crystalline silicon substrate the atom to atom spacing is 0:222 nm, meaning that
there are between 4 to 5 silicon atoms in one nanometer.
xxv
xxvi Explanation of Atomic-Scale Terms

Molecule: a molecule is the smallest part of a substance that still incorporates the
chemical properties of the substance. It is built from an electrically neutral group of
atoms, which are bound to each other in a fixed order. The mass of a molecule is the
sum of the masses of the individual atoms, from which it is built. A simple hydrogen
molecule (H2 ), for example, only consists of two hydrogen atoms which are bound
by one electron pair. A water molecule (H2 O) consists of two light hydrogen atoms
and one (about 16 times heavier) oxygen atom.

Ion: an ion is an electrically charged atom or molecule or other group of bound

atoms, created by the removal or addition of electrons by radiation effects or
chemical reactions. It can be positively or negatively charged by, respectively, the
shortage or surplus of one or more electrons.
List of Physical Constants

0 D 8:85  1012 F/m

ox D 4 for silicon dioxide
si D 11:7
f D 0:5 V for silicon substrate
k D 1:4  1023 Joule/K
q D 1:6  1019 Coulomb

Bandgap for Si: 1.12 eV

Bandgap for SiO2 : 9 eV

xxvii
Contents

1 Basic Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 1
1.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 1
1.2 The Field-Effect Principle . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 1
1.3 The Inversion-Layer MOS Transistor . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 4
1.3.1 The Metal-Oxide-Semiconductor (MOS) Capacitor . . . . 9
1.3.2 The Inversion-Layer MOS Transistor .. . . . . . . . . . . . . . . . . . . 11
1.4 Derivation of Simple MOS Formulae . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 18
1.5 The Back-Bias Effect (Back-Gate Effect, Body Effect)
and the Effect of Forward-Bias . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 21
1.6 Factors Which Characterise the Behaviour of the MOS
Transistor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 25
1.7 Different Types of MOS Transistors . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 26
1.8 Parasitic MOS Transistors . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 28
1.9 MOS Transistor Symbols . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 29
1.10 Capacitances in MOS Structures . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 30
1.11 Conclusions .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 38
1.12 Exercises .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 39
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 43
2 Geometrical-, Physical- and Field-Scaling Impact on MOS
Transistor Behaviour .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 45
2.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 45
2.2 The Zero Field Mobility . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 45
2.3 Carrier Mobility Reduction . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 46
2.3.1 Vertical and Lateral Field Carrier Mobility Reduction . . 47
2.3.2 Stress-Induced Carrier Mobility Effects . . . . . . . . . . . . . . . . . 50
2.4 Channel Length Modulation . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 51
2.5 Short- and Narrow-Channel Effects . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 53
2.5.1 Short-Channel Effects .. . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 53
2.5.2 Narrow-Channel Effect . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 55
2.6 Temperature Influence on Carrier Mobility and Threshold
Voltage .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 57

xxix
xxx Contents

2.7 MOS Transistor Leakage Mechanisms .. . . . . . . . . .. . . . . . . . . . . . . . . . . . . 59

2.7.1 Weak-Inversion (Subthreshold) Behaviour
of the MOS Transistor . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 60
2.7.2 Gate-Oxide Tunnelling .. . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 62
2.7.3 Reverse-Bias Junction Leakage . . . . . . .. . . . . . . . . . . . . . . . . . . 64
2.7.4 Gate-Induced Drain Leakage (GIDL) .. . . . . . . . . . . . . . . . . . . 65
2.7.5 Hot-Carrier Injection and Impact Ionisation . . . . . . . . . . . . . 66
2.7.6 Overall Leakage Interactions and Considerations . . . . . . . 66
2.8 MOS Transistor Models and Simulation . . . . . . . . .. . . . . . . . . . . . . . . . . . . 68
2.8.1 Worst-Case (Slow), Typical and Best-Case (Fast)
Process Parameters and Operating Conditions.. . . . . . . . . . 69
2.9 Conclusions .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 70
2.10 Exercises .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 71
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 71
3 Manufacture of MOS Devices .. . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 73
3.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 73
3.2 Different Substrates (Wafers) as Starting Material . . . . . . . . . . . . . . . . . 74
3.2.1 Wafer Sizes . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 75
3.2.2 Standard CMOS Epi . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 76
3.2.3 Crystalline Orientation of the Silicon Wafer . . . . . . . . . . . . . 78
3.2.4 Silicon-on-Insulator (SOI) . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 79
3.3 Lithography in MOS Processes . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 83
3.3.1 Lithography Basics . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 83
3.3.2 Lithographic Extensions Beyond 30 nm . . . . . . . . . . . . . . . . . 95
3.3.3 Next Generation Lithography.. . . . . . . . .. . . . . . . . . . . . . . . . . . . 101
3.3.4 Mask Cost Reduction Techniques for
Low-Volume Production .. . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 105
3.3.5 Pattern Imaging.. . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 107
3.4 Oxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 108
3.5 Deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 112
3.6 Etching .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 117
3.7 Diffusion and Ion Implantation .. . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 120
3.7.1 Diffusion .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 120
3.7.2 Ion Implantation .. . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 121
3.8 Planarisation .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 123
3.9 Basic MOS Technologies . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 128
3.9.1 The Basic Silicon-Gate nMOS Process . . . . . . . . . . . . . . . . . . 128
3.9.2 The Basic Complementary MOS (CMOS) Process . . . . . 131
3.9.3 An Advanced Nanometer CMOS Process . . . . . . . . . . . . . . . 133
3.9.4 CMOS Technologies Beyond 45 nm . .. . . . . . . . . . . . . . . . . . . 141
3.10 Conclusions .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 155
3.11 Exercises .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 156
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 157
Contents xxxi

4 CMOS Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 161

4.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 161
4.2 The Basic nMOS Inverter .. . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 162
4.2.1 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 162
4.2.2 The DC Behaviour . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 163
4.2.3 Comparison of the Different nMOS Inverters . . . . . . . . . . . 170
4.2.4 Transforming a Logic Function into an nMOS
Transistor Circuit .. . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 171
4.3 Electrical Design of CMOS Circuits . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 173
4.3.1 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 173
4.3.2 The CMOS Inverter . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 174
4.4 Digital CMOS Circuits .. . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 187
4.4.1 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 187
4.4.2 Static CMOS Circuits . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 187
4.4.3 Clocked Static CMOS Circuits . . . . . . . .. . . . . . . . . . . . . . . . . . . 192
4.4.4 Dynamic CMOS Circuits . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 195
4.4.5 Other Types of CMOS Circuits . . . . . . . .. . . . . . . . . . . . . . . . . . . 200
4.4.6 Choosing a CMOS Implementation .. .. . . . . . . . . . . . . . . . . . . 200
4.4.7 Clocking Strategies. . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 201
4.5 CMOS Input and Output (I/O) Circuits . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 202
4.5.1 CMOS Input Circuits . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 202
4.5.2 CMOS Output Buffers (Drivers) . . . . . .. . . . . . . . . . . . . . . . . . . 203
4.6 The Layout Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 204
4.6.1 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 204
4.6.2 Layout Design Rules . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 205
4.6.3 Stick Diagram . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 208
4.6.4 Example of the Layout Procedure .. . . .. . . . . . . . . . . . . . . . . . . 211
4.6.5 Guidelines for Layout Design . . . . . . . . .. . . . . . . . . . . . . . . . . . . 214
4.7 Libraries and Library Design . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 215
4.8 FinFET Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 218
4.9 Conclusions .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 221
4.10 Exercises .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 222
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 224
5 Special Circuits, Devices and Technologies . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 227
5.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 227
5.2 CCD and CMOS Image Sensors.. . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 227
5.2.1 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 227
5.2.2 Basic CCD Operation . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 228
5.2.3 CMOS Image Sensors.. . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 231
5.3 BICMOS Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 233
5.3.1 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 233
5.3.2 BICMOS Technology .. . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 234
5.3.3 BICMOS Characteristics. . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 235
xxxii Contents

5.3.4 BICMOS Circuit Performance . . . . . . . .. . . . . . . . . . . . . . . . . . . 237

5.3.5 Future Expectations and Market Trends . . . . . . . . . . . . . . . . . 239
5.4 Power MOSFETs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 239
5.4.1 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 239
5.4.2 Technology and Operation .. . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 240
5.4.3 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 241
5.5 Bipolar-CMOS-DMOS (BCD) Processes . . . . . . . .. . . . . . . . . . . . . . . . . . . 243
5.6 Conclusions .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 246
5.7 Exercises .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 246
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 247
6 Memories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 249
6.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 249
6.2 Serial Memories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 252
6.3 Content-Addressable Memories (CAM) . . . . . . . . .. . . . . . . . . . . . . . . . . . . 253
6.4 Random-Access Memories (RAM) . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 253
6.4.1 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 253
6.4.2 Static RAMs (SRAM) . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 253
6.4.3 Dynamic RAMs (DRAM) . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 264
6.4.4 High-Performance DRAMs. . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 275
6.4.5 Single- and Dual Port Memories . . . . . .. . . . . . . . . . . . . . . . . . . 280
6.4.6 Error Sensitivity .. . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 281
6.5 Non-volatile Memories . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 281
6.5.1 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 281
6.5.2 Read-Only Memories (ROM) . . . . . . . . .. . . . . . . . . . . . . . . . . . . 282
6.5.3 Programmable Read-Only Memories .. . . . . . . . . . . . . . . . . . . 285
6.5.4 EEPROMs and Flash Memories .. . . . . .. . . . . . . . . . . . . . . . . . . 287
6.5.5 Non-volatile RAM (NVRAM) . . . . . . . .. . . . . . . . . . . . . . . . . . . 304
6.5.6 BRAM (Battery RAM) . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 304
6.5.7 FRAM, MRAM, PRAM (PCM) and RRAM . . . . . . . . . . . . 304
6.6 Embedded Memories.. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 308
6.6.1 Redundancy . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 312
6.7 Classification of the Various Memories . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 314
6.8 Conclusions .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 314
6.9 Exercises .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 316
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 317
7 Very Large Scale Integration (VLSI) and ASICs . . . .. . . . . . . . . . . . . . . . . . . 321
7.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 321
7.2 Digital ICs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 323
7.3 Abstraction Levels for VLSI .. . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 327
7.3.1 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 327
7.3.2 System Level . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 331
7.3.3 Functional Level.. . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 333
7.3.4 RTL Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 334
7.3.5 Logic-Gate Level .. . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 336
Another random document with
no related content on Scribd:
beginning with Frederick II., and ends with a proud assertion that he
will speak the whole truth in spite of the powerful barons whom he
may offend. So here, in the Valley of the Princes, where those are
detained who neglected some peculiarly lofty mission, or postponed
their spiritual welfare to worldly and political care, Sordello,
beginning with the Emperor-elect, Rudolph of Hapsburg (Purg. vii.
94), points out the descendants or successors of those whom he had
rebuked in the other life. Here, singing together to the Queen of
Mercy, the deadliest foes sit side by side, consoling each other;
Rudolph of Hapsburg with Ottocar of Bohemia, Charles of Anjou with
Peter of Aragon; a motive found previously in the vision of Tundal,
where, however, the kings are naturally Irish. On Henry of England
Sordello had been more severe when he lived. After sunset, in the
light of three brighter stars, that symbolise the three theological
virtues, Dante has pleasant talk with Nino Visconti and Currado
Malaspina (Purg. viii.). And, as evening closes in, two golden-haired
Angels, green-clad and green-winged, the Angels of Hope with the
flaming but blunted swords of justice tempered with mercy, defend
the noble souls from the assault of an evil serpent. In the literal
sense, this episode (which seems a relic from earlier mediaeval
visions) may imply that souls in Purgatory have not the intrinsic
impossibility of sinning that is possessed by the blessed of Paradise,
but are kept absolutely free from any sin by the Divine Providence. In
the allegorical sense, the meaning clearly is that the way to moral
and intellectual freedom is a hard one, and temptations to fall back in
despair are many. The tempter would draw man back from regaining
the Earthly Paradise, from which he has once caused his expulsion.
The Mystic Eagle and the Gate of Purgatory.—Just before
the dawn Dante dreams of a golden eagle snatching him up to the
sphere of fire, and, waking when the sun is more than two hours
high, finds that Lucia has brought him to the Gate of Purgatory.
Mystically, the eagle seems to represent the poet’s own spirit,
dreaming that he can soar unaided to the very outskirts of Paradise;
but he wakes to realise that Divine grace indicates the preliminary
stage of purification. The gate of St. Peter with its three steps, of
white marble, exactly mirroring the whole man, of darkest purple
cracked in the figure of the Cross, of flaming red porphyry,
represents the Sacrament of Penance with its three parts: Contrition,
Confession, Satisfaction based upon the love of God. The mournfully
robed Angel of Obedience seated on the rock of diamond, with
dazzling face and flashing sword, is the confessor. His silver and
gold keys, of judgment and absolution, open the gate to Dante; the
seven P’s traced by his sword on the poet’s forehead are to be
effaced one by one in his ascent (Purg. ix.).
Moral Topography.—Within the gate is Purgatory proper with its
seven terraces, each devoted to the purgation of one of the seven
capital sins, “out of which other vices spring, especially in the way of
final causation” (Aquinas). Whereas in the Inferno sin was
considered in its manifold and multiform effects, in the Purgatorio it is
regarded in its causes, and all referred to disordered love. The
formal element, the aversion from the imperishable good, which is
the essence of Hell, has been forgiven; the material element, the
conversion to the good that perishes, the disordered love, is now to
be purged from the soul. In the allegorical or moral sense, since
love, as Aquinas says, is “the ultimate cause of the true activities of
every agent,” it is clear that man’s first duty in life is to set love in
order; and, indeed, the whole moral basis of Dante’s Purgatory rests
upon the definition of St. Augustine that virtue is ordo amoris, “the
ordering of love.” In the first three terraces, sins of the spirit are
expiated; in the fourth terrace, sloth, which is both spiritual and
carnal; in the fifth, sixth, seventh terraces, sins of the flesh. This
purgation, which involves both pain of loss for a time and
punishment of sense, is effected by turning with fervent love to God
and detesting what hinders union with Him. Therefore, at the
beginning of each terrace, examples are seen or heard of virtue
contrary to the sin, in order to excite the suffering souls to extirpate
its very roots; and, at the end, examples of its result or punishment
(the “bit and bridle”). These examples are chosen with characteristic
Dantesque impartiality alike from Scripture and legend or mythology;
but, in each case, an example from the life of the Blessed Virgin is
opposed to each capital sin. At the end of each terrace stands an
Angel—personification of one of the virtues opposed to the sins or
vices. These seven Angels in their successive apparitions are
among the divinest things of beauty in the sacred poem. It is only
when sin is completely purged away that man can contemplate the
exceeding beauty, the “awful loveliness” of the contrary virtue.
First Terrace.—Steep and narrow is the path up to the first
terrace, where Pride is purged away (Purg. x). Carved upon the
mountain side are fair white marble images of wondrous beauty,
setting forth great examples of Humility, alike in “them of low degree”
(Mary at the Annunciation) and in “the mighty” (David and Trajan,
rulers respectively of the chosen people of the two dispensations,
the Jews and the Romans). Wearily and painfully the souls of the
proud pass round, pressed down by terrible weights, reciting a
paraphrase of the Lord’s Prayer, for themselves and those they have
left on earth. And seldom has the Catholic doctrine of prayer for the
dead been more winningly set forth than in Dante’s comment (xi. 31-
36). A partaker in some degree of their punishment, Dante, all
bowed down, goes with these souls; he speaks with Omberto
Aldobrandesco, who is expiating pride of birth, and Oderisi of
Gubbio, the miniaturist, who is purifying his soul from pride of
intellect. The latter points out the great Ghibelline burgher statesman
of Siena, Provenzano Salvani, expiating pride of dominion—the sin
which turned so many an Italian patriot of the Middle Ages into a
tyrant. Figured upon the pavement below their feet are examples of
Pride’s punishment, like the designs on the pavement of the Duomo
of Siena (Purg. xii.). Noon has passed when the Angel of Humility
shows the way up to the next terrace, and with the waving of his
wing removes the first P from Dante’s forehead. “Blessed are the
poor in spirit,” celestial voices sing, as, with almost all weariness
gone since Pride is expiated, Dante ascends the steep way.
Second Terrace.—In the second and narrower circle Envy is
purged. Examples of charity, “courteous invitations to the table of
Love,” are cited by invisible spirits flying past. The envious, clothed
in haircloth, lean helplessly shoulder to shoulder against the rock,
their eyelids sewn up with iron stitching. Sapia of Siena, the
kinswoman of Provenzano Salvani, at whose fall and the defeat of
her countrymen she rejoiced, tells her history in lines of singular
beauty (Purg. xiii.). Guido del Duca denounces the evil dispositions
of the inhabitants of Tuscany, and bewails the degeneracy of the
noble houses with the consequent decay of chivalry in his own
province of Romagna; envious on earth of prosperity of others, these
souls mourn now for its decline (xiv.). Like peals of thunder the cries
of spirits follow each other in citing Envy’s punishment. As they go
towards the sunset, the dazzling Angel of Fraternal Love removes
the mark of Envy. “Blessed are the merciful,” “Rejoice thou that
conquerest.” As they mount Virgil expounds the difference between
material goods, which are diminished by sharing and beget envy,
and the infinite good of Paradise, where love increases with every
soul that enters into the joy of the Lord, and its communication is
measured only by the charity of each soul that is made its mirror
(Purg. xv.).
Third Terrace.—On reaching the third terrace where Anger is
purged, Dante sees examples of meekness and forgiveness in
vision. From the black, pungent, and tormenting smoke which
envelopes the souls of the once wrathful, who now call upon the
Lamb of God for peace and mercy, the Lombard Marco reconciles
Free Will with stellar influence, and ascribes the evil condition of Italy
and the world to the neglect of law, the confusion of the spiritual and
temporal power, and the papal usurpation of imperial rights (Purg.
xvi.). In this terrace Dante again partakes of the pains of the penitent
souls. As the sun is setting, he issues from the dark mist. A most
significant passage on the power of the imagination to form images
not derived from the senses (xvii. 13-18) introduces the visions of
Anger’s punishment, from which the poet is roused by the dazzling
splendour of the Angel of Peace or Meekness, who fans away the
third P and shows the way up: “Blessed are the peacemakers who
are without evil wrath.”
Fourth Terrace.—The stars are appearing as they reach the
fourth terrace, where souls are purged from Sloth. We saw that, in
the Inferno, the Aristotelian division of things to be morally shunned
was discussed, and the ethical structure of the first canticle
expounded, in the circle intermediate between Incontinence and
Malice (Inf. xi.); so, in the Purgatorio, a compulsory pause in the
terrace intermediate between sins of spirit and sins of flesh is
selected by Virgil for his great discourse upon Love, on which is
based the moral system of the second realm (Purg. xvii. 91-139,
xviii. 13-75). It is practically a sermon on the text of Jacopone da
Todi, Ordena questo amore, tu che m’ami, “Set this love in order,
thou that lovest me”; since in rational beings disordered love
produces the seven capital vices. Pride, Envy, Anger are regarded
as distorted love; Sloth as defective love; Avarice, Gluttony, Luxury
as excessive love. Love is the golden net whereby God draws back
to Himself all creatures that He has made, whether inanimate,
sensitive, or rational—by the tendencies or inclinations He has given
them to make them seek the end for which they are ordered and
disposed, according to the Eternal Law. Rational beings alone have
Free Will, by which man merits or demerits from the Divine Justice,
according as he inclines to good or evil loves. Love’s tendency to
good is the precious material upon which Free Will acts like the
craftsman’s hand, to fashion a satyr’s mask or a crucifix.
At the end of this discourse, the slothful rush by at full speed in the
moonlight, so full of longing to lose no time through too little love,
that the Abbot of San Zeno cannot stop while he answers Virgil’s
question; those in front cry out examples of alacrity in Mary and
Caesar; those behind chant Sloth’s punishment in the chosen people
of the Old Testament and the Trojan ancestors of the Romans.
The Siren and the Angel of Zeal.—Before the dawn of the third
day in Purgatory, Dante has in his sleep a marvellous dream of the
Siren (sensual seduction, concupiscence of the flesh), from which he
is delivered by a holy and alert lady who calls upon Virgil (prevenient
grace, or the wisdom and prudence of Proverbs vii.). The Siren is the
dream-prelude to the purgation of sins of the flesh, as the Eagle had
been to that of sins of the spirit. The sun has risen; and the Angel of
Zeal (or of Spiritual Joy) cancels the fourth P and shows the way up
to the next terrace. “Blessed are they that mourn, for they shall have
their souls wed to consolation.” Sloth is a heaviness and sadness
which weighs down the soul, a sadness at spiritual good, to be
fought by thinking on spiritual things. Most fitly then do the wings of
the Angel of Zeal point upwards, and his words tell of a nobler
sorrow, a mourning which shall be followed by Divine consolation
(Purg. xix.).
 Fifth Terrace.—In the fifth terrace, the avaricious and prodigal,
whose souls on earth cleaved to the dust, lie face downwards to
earth; unable to move hand or foot until the sin of Covetousness is
purged away, the sin which, according to Aquinas, “although not
absolutely the greatest of sins, yet has in some sense a greater
deformity than the rest, since by it the human heart is subjected even
to external things.” Pope Adrian V. tells the story of his tardy
conversion, and has tender words for his niece Alagia, the wife of
Moroello Malaspina (Purg. xix.). It is a companion episode to that of
Nicholas III. in the corresponding canto of the Inferno. In this circle
the souls themselves cry out the examples and warnings, by day and
night respectively. The soul of Hugh Capet, “the root of the evil plant
which overshadows all the Christian earth,” pours forth bitter
sarcasm and scathing invective upon all the royal house of France,
the great Guelf power that opposed the Empire, oppressed Italy, and
wrought scandal in the Church. A monument of poetic infamy is
especially raised to Philip the Fair and the three Carlos; and there
are few more glorious examples of Christian magnanimity than the
burning words in which Dante, distinguishing the man from the office,
brands the sacrilege of Anagni, the outrage committed upon him
whom the poet held as his own deadliest foe, and yet the unworthy
Vicar of Christ. Nowhere else, save in the reference to the Jubilee
(Purg. ii. 98, 99), does Dante treat Boniface as lawful pope (cf. Inf.
xix. 52-57; Par. ix. 142, xxvii. 22-24). It has been thought that Canto
xx. was composed while the Church was ostensibly supporting the
policy of Henry VII.; before attacking the Templars, the French king
had endeavoured to renew the outrage of Anagni by inducing Pope
Clement to condemn the memory of Boniface. With a mighty
earthquake, a universal chorus of Gloria in excelsis from the
suffering souls, the poet Statius is liberated, and joins Dante and
Virgil (Purg. xxi.). He explains how the pains of Purgatory are
voluntarily endured, since, against the hypothetic or absolute will
with which they desire the bliss of Paradise, the souls suffer these
purifying pains with the conditional or actual will, the same inclination
or impulse or desire (talento) which they formerly had to sin. Thus it
is free will itself that imposes the purgatorial process, and that alone
shows the soul when purification is complete. The delicious scene of
the recognition of Virgil by Statius is full of that peculiarly tender
Dantesque playfulness that informs the two Eclogues; Dante’s
affectionate humour in dealing with those he loved is one of the most
attractive aspects of his character, and one perhaps too often
missed.
Sixth Terrace.—The Angel of Justice has removed the fifth P
from Dante’s forehead, opposing in his song the thirst of justice to
that of gain. As they mount, Statius explains to Virgil how he was
converted from prodigality by a line in the Aeneid, and led to
Christianity by the fourth Eclogue (Purg. xxii.). The conversion of a
pagan to Christianity through reading Virgil occurs in a story told by
Vincent of Beauvais; Dante was probably influenced in applying this
to Statius, representing him as a secret convert to the true faith, by
his study of the Thebaid; for there, in the last book, Statius describes
the Altar of Mercy at Athens in language which harmonises with the
words of Christ in the Gospels and the address of his own
contemporary, St. Paul, to the Athenians in the Acts. The poets
pursue their way with greater confidence now that Statius is with
them, and reach the sixth terrace, where unseen spirits cry out
examples of temperance from the tree beneath which drunkenness
and gluttony are purged. The spirits, terribly wasted, suffer intense
torments of hunger and thirst in the presence of most tempting food
and drink; but the sanctifying pain is a solace, desired even as Christ
willed to die for man. With the soul of Forese Donati, Dante holds
loving converse; the memory of their dissolute lives together is still
grievous; the poet makes amends for his old slander of Forese’s wife
Nella, by the tender lines now placed upon her husband’s lips (Purg.
xxiii. 85-93). Forese darkly foretells the death of Corso Donati, which
appears to be the latest event in Florentine history mentioned in the
poem (xxiv. 82-90). Whatever the friendship of these two had been
on earth, it was fair and lovely indeed on the Mount of Purgation.
Amongst many others are Pope Martin IV. and the poet
Bonagiunta of Lucca, whose talk with Dante upon the dolce stil
nuovo, the “sweet new style,” is one of the landmarks for the student
of poetry (Purg. xxiv. 49-60). Dante’s famous definition of his own
position expresses, in another form, the truth that all great poetry is
the “transfigured life” of its author:[35] “I am one who, when Love
inspires me, note, and give utterance in that fashion which he
dictates within.” It is already anticipated in the prose passage
prefixed to the Donne che avete in the Vita Nuova (xix.), and
completes the conception of poetry set forth in the De Vulgari
Eloquentia.
The Seventh Terrace.—Passing another tree, a shoot from the
tree of knowledge, beneath which the purging pangs are renewed,
and from whose branches spirit voices proclaim examples of
gluttony’s punishment, they are summoned upwards by the glowing
and dazzling Angel of Abstinence, fragrant with grass and flowers as
the air of May. As they ascend the narrow stairs towards the last
terrace, Statius explains the generation of the body and the infusion
of the rational soul, which exists, after the body’s death, invested
with an aerial body as a shade (Purg. xxv. 31 et seq.). Apparently it
is because revelation has some voice in these high matters that the
Christian Statius gives Dante this exposition, instead of Virgil, and at
the latter’s request; until the seventh terrace is reached, where
sensual passion is expiated in the bosom of the great burning.
Singing to the God of Supreme Clemency, crying aloud examples of
chastity or of lust’s punishment, two bands of souls, divided
according to the nature of their sin, pass through the fire in opposite
ways (Purg. xxvi.). Here is Guido Guinizelli of Bologna, father of the
poets of the dolce stil nuovo, whom Dante gazes upon in rapt
admiration, and addresses with impassioned love and worship. But
Guinizelli—with that humility which is so characteristically Dante’s
own—indicates as miglior fabbro del parlar materno, a “better
craftsman of his mother-tongue,” Arnaut Daniel, the cunning
Provençal song-smith, who invented the sestina, and whose metrical
skill and originality won for him a higher place in the estimation of the
poet of the rime pietrose than modern students of the troubadours
are usually disposed to concede.
The Purging Fire.—At sunset the Angel of Purity, singing
“Blessed are the clean of heart,” bids the poets pass through the
flames that lie between them and the last stairway—the purging fire
that is the wall between Dante and Beatrice. Dante endures the
“burning without measure”; and they reach the ascent, greeted by
dazzling light and celestial strains of Venite benedicti Patris mei. The
Cherubims with the flaming sword, “turning every way to keep the
way of the tree of life” (Gen. iii. 24), are thus welcoming man’s
restoration to the Garden of Eden, as the serpent had endeavoured
to impede it in the Valley of the Princes. Now it is a delight to mount;
but night comes on, and Dante, watched over by Statius and Virgil,
falls asleep on the stairs (Purg. xxvii.).
Leah and Liberty.—Just before dawn, prelude to the new day, he
dreams of Leah, a young and lovely lady gathering flowers in a
meadow. The theologians took Leah as type of the active life, and
Rachel, her sister, of the contemplative; a symbolism to which
Richard of St. Victor gave a more mystical colour, by interpreting
Leah as “affection inflamed by divine inspiration, composing itself to
the norm of justice.” Leah may then represent the affection, thus
inflamed and ordered, which is the perfection of the active life. At
sunrise the topmost stair of Purgatory is reached, and Virgil, who can
himself discern no further, resigns his guidance at the entrance to
the Garden of Eden. Dante’s judgment has been made free, right,
and whole; per ch’ io te sovra te corono e mitrio, “wherefore I crown
and mitre thee over thyself” (xxvii. 142). It has been supposed that
Virgil is here resigning to Dante the crown and mitre of the Emperor;
mitratus et coronatus was the expression used for the coronation of
an Emperor when the Pope placed upon his head a mitre and a
crown, which afterwards were united in the mitred crown, as seen in
the great fresco at Santa Maria Novella. Others refer the crown to
temporal or imperial authority, and the mitre to spiritual or
ecclesiastical; for (Mon. iii. 4) “if man had remained in the state of
innocence in which he was made by God, he would have had no
need of such directive regimens,” which are “remedial against the
infirmity of sin.” Dante, purified from sin, has regained this state of
innocence, and has attained that liberty through which “we have our
felicity here as men and our felicity elsewhere as Gods” (Mon. i. 12).
In any case, Virgil is confirming the freedom which Dante has sought
and gained by the passage through Purgatory.
 The Earthly Paradise and Matelda.—The Earthly Paradise
represents “blessedness of this life, which consists in the exercise of
man’s natural powers” (Mon. iii. 16). This blessedness is found in the
twofold exercise of the mind: the practical, which “consists in
ourselves working virtuously, that is, in integrity, with prudence, with
temperance, with fortitude, and with justice”; and the speculative,
which consists “in considering the works of God and of nature”
(Conv. iv. 22). In this Earthly Paradise, the music of whose birds and
trees has surely passed into the wonderful six cantos that close the
Purgatorio, Dante meets, amidst the flowers on Lethe’s banks, the
glorified realisation of the Leah of his dream (Purg. xxviii.). She has
been taken as symbolising the glorified active life in the state of
recovered Eden, realising in the Church of Christ what Leah had
dimly prefigured in the Old Testament; the active Christian life;
innocentia bonorum operum, the virtuous use of earthly things,
directly ordered to the love of our neighbour; the temporal felicity of
the Earthly Paradise. Since the purgatorial process is the freeing of
the soul from disordered love, we may follow Richard’s interpretation
of Leah, and take her as representing love rightly ordered and
inflamed by divine inspiration. Presently she is called Matelda (xxxiii.
119), and it is probable that she is the idealised presentment of a
real person. All the earliest commentators, excepting the Ottimo,
identify her with the great Countess of Tuscany, in support of which
view might be urged the historical work of the Countess in the revival
of the study of Roman Law at Bologna—Roman Law being, for
Dante, the secular counterpart of the “perfect law of liberty.” Some
modern commentators prefer to seek her prototype in one or other of
the ladies of Vita Nuova; for instance, in that lady of very sweet
speech who had rebuked Dante at the crisis of his “new life.” Others
have attempted to identify her with Mechthild of Magdeburg or
Mechthild of Hackeborn, two German mystical writers of the latter
part of the thirteenth century whose works show occasional
analogies with the Commedia. It may be observed that her
counterpart, as Rachel to Leah, is not Beatrice, as sometimes
supposed, but St. Bernard, in the closing cantos of the Paradiso.
Matelda explains her joyous aspect by referring Dante to the Psalm
Delectasti (Ps. 92, 91 Vulgate), and her discourse of Eden and its
rivers (realising the Golden Age sung by the classical poets)
communicates to Virgil and Statius her own celestial joy: “Thou has
given me, O Lord, a delight in what Thou hast made: in the works of
Thy hands I shall rejoice.” She points out to Dante’s gaze the
wondrous pageant, which astonishes Virgil as much as his pupil, the
mystical procession that represents the triumphal march of the
Church (Purg. xxix.).
The Pageant of the Church.—With brilliant light and ineffable
melody, the triumph advances: “I saw the holy city, the new
Jerusalem, coming down out of heaven from God, prepared as a
bride adorned for her husband” (Rev. xxi. 2). Headed by seven
candlesticks of gold as standards, followed by the twenty-four elders,
white-robed and crowned with lilies, singing Mary’s praises; between
the four living creatures of Ezekiel and St. John, crowned with green,
comes a triumphal chariot, more glorious than the sun, upon two
wheels; drawn by a Griffin, half lion and half eagle, whose golden
wings stretch up far out of sight, through the seven luminous bands
that form the processional canopy. By the right wheel dance three
maidens, symbolic of the theological virtues; by the left wheel dance
four, who represent the cardinal virtues, following the measure of
Prudence, as the others take their step from the song of Charity. The
seven candlesticks are the gifts of the Holy Spirit; the twenty-four
elders, either the patriarchs and prophets, or the books of the Old
Testament; the four living creatures, the four Evangelists, or their
four Gospels; the Griffin, Christ Himself in His Human and Divine
Natures. Lastly, follow seven more elders, white-robed but crowned
with flaming red flowers; a physician, and one with shining sword;
four of humble appearance; an old man “sleeping with face alert.”
According to Benvenuto da Imola, these represent St. Peter (who
had intrusted to him the power of healing souls) and St. Paul, the
four great Latin doctors, and St. Bernard. More usually they are
regarded as personifying the books of the New Testament—the Acts,
St. Paul’s Epistles, the Epistles of St. Peter, James, John, and Jude,
the Apocalypse or Revelation of St. John. Upon the chariot, amidst a
hundred Angels singing and scattering flowers, Beatrice appears,
clad in the mystical colours, red, white, green, crowned with the olive
of wisdom and of peace over her snow-white veil. And, at the advent
of the Wisdom divinely revealed to man, Virgil silently vanishes; he
has tasted of the delights of the Earthly Paradise, has witnessed the
triumph of the Church from which he is for ever cut off, the Faith he
never knew, and has gone back to his mournful dwelling-place (Purg.
xxx.).
Beatrice and Dante.—The precise significance of the
reproaches which Beatrice pours upon Dante for his mode of life
after her death, with the poet’s own bitter shame and intense
repentance (xxx., xxxi.), depends upon the view taken of his
character and the nature of the wanderings represented in the dark
wood. That these aberrations were mainly philosophical and
intellectual, as sometimes supposed, appears highly improbable. We
would regard Dante’s confession here as one of his most personal
utterances, and hold that the cherubically inspired singer of
righteousness is deliberately casting aside the allegorical veil which,
in the Convivio, he had attempted to throw over the things in the past
which still severed him from the ideal life when he wrote: “I fear the
infamy of having followed such great passion.” It is a personal
episode, in which Beatrice is the woman loved and to whose
memory the poet has been unfaithful, standing out clearly from the
allegorical mystery by which it is surrounded and in which it is set.
After Matelda has drawn Dante through Lethe, the four cardinal
virtues, which “perfect the intellect and appetite of man according to
the capacity of human nature,” lead him to the breast of the mystic
Griffin; and, in response to the song of the three theological virtues,
which perfect man supernaturally, Beatrice at last unveils her
countenance to his gaze: “O splendour of living light eternal.”
Concluding Allegories of the “Purgatorio.”—The allegory is
resumed. In the light of this revelation, now that he is purified and
free from sin, Dante beholds a vision of the Church and Empire
(Purg. xxxii.). That glorious procession had first presented an ideal of
the Church as Divine Providence intended it to be, before it became
the vessel that the serpent of simony broke; the Bride that the Divine
Spouse ordained for the guidance of the world. Such being the ideal,
Dante beholds in a series of allegorical visions its history, in
conjunction with the Empire, from the first coming to Rome down to
the transference of the papal chair to Avignon. The great procession
moves on through the divine forest, the Griffin still drawing the
chariot with Beatrice seated upon it; Matelda with Dante and Statius
following after the right wheel. Even as the divine origin of the
Church has been seen in the triumphal car, so now the divine origin
of the Empire is indicated in the desolate and despoiled tree which
they reach. The tree of knowledge of good and evil, since the
prohibition to eat of that tree was the beginning of law and the duty
of obedience, represents Natural Law or Natural Justice, what Dante
calls ius; which “in things is nought else than the similitude of the
divine will” (Mon. ii. 2). The expression of this natural justice and the
means for its effectuation in human society is Law, which Dante
identifies with the Empire, and thus the tree becomes the symbol of
the Empire and of the obedience due to it. The tree is destitute of
flowers and foliage till the Griffin comes to it, who plucks nothing
from it: “Thus is preserved the seed of all justice” (Purg. xxxii. 48; cf.
our Lord’s words to St. John, Matt. iii. 15). Justice can alone be
fulfilled when the Church follows this example of her Divine Founder,
and usurps none of the temporal rights of the Empire. After the
chariot has been bound to the tree, the previously bare plant breaks
out into purple leaves and flowers. The Griffin and his train return to
Heaven, leaving Beatrice to guard the chariot of the Church, seated
beneath the shadow of the Imperial Tree, upon its root, which is
Rome. In a new series of visions Dante beholds the sequel; he sees
the conflict of the past, contemplates the corruption of the present,
hearkens to the hope of the future. The persecution of the Church by
the early Roman Emperors is followed by the inroad of the first
heresies; and the donation of Constantine by the rising of the dragon
of schism or simony. By more assumption of secular power and
dignities, the chariot becomes monstrously transformed, and
shamelessly usurped by the harlot, who represents the corrupt
ecclesiastical authority enthroned in the place of Revelation, a false
and degraded theology based upon the Decretals instead of the true
divine science of the Scripture and the Fathers. By her side a giant
appears who, after alternate caressing and scourging of the usurper,
unbinds the transfigured chariot from the tree, and drags it away
through the forest—symbolical of the interference of the royal house
of France, ending in the transference of the Papacy from Rome to
Avignon.
A Deliverer Announced.—But to the mournful psalm that the
maidens around her raise, Deus venerunt gentes, Beatrice answers
in words of hope; “a little while,” and the spiritual guide shall rise
again from the black tomb of Avignon. And, as they move on, she
utters to Dante a further prophecy (Purg. xxxiii.). “The vessel that the
serpent broke was and is not,” so completely has corruption and
simony degraded the chariot of the Bride of Christ. But vengeance
shall fall upon the guilty parties, and the eagle shall not for ever be
without an heir; for already a favourable disposition of the stars is at
hand, under which a messenger of God shall come, who shall slay
the harlot and the giant. It is probably the same event as the coming
of the Veltro. Dante is to repeat her words “to those that live the life
which is a running to death,” and not to conceal what he has seen of
the tree. Apparently (Purg. xxxiii. 58-72) he is to make manifest that
the Empire is of divine origin, and to recognise that the precept given
by God to our first parents corresponds now with the duty and
obedience man owes to the Empire. The law under which Adam
lived was the prohibition to eat of the tree; the law under which his
descendants, the commonwealth of the human race, live is the
Empire. As Parodi puts it, it is not a new sense superimposed upon
the first; “it is simply the same single meaning, the historical
circumstances alone appearing changed.” The sin of Adam is
repeated when the Empire is usurped of its rights or its authority
attacked, for God created it holy for the purpose of leading man to
temporal felicity—the goal, here and now, of the human race.[36]
Lethe and Eunoë.—At noon they come to where the rivers of
Lethe and Eunoë issue from one mystical fountain, the fountain of
the grace of God. Here Beatrice refers Dante to Matelda, who leads
him and Statius to drink of Eunoë, which quickens dead virtue and
restores memory of every good deed in those who have first been
bathed in Lethe, which takes away the memory of sin. According to
St. Thomas Aquinas (Summa, iii. 89, 5), works done in charity,
although in a sense dead through sin, are brought to life through
penance. Through repentance they regain their efficacy of leading
him who did them into eternal life. Therefore Dante writes: “I returned
from the most holy stream, remade even as young trees renewed
with new foliage, pure and disposed to ascend to the stars.”

4. The “Paradiso”
Structure.—Dante’s Paradise consists of the nine moving
heavens, according to Ptolemaic astronomy, crowned by the tenth
motionless and divinest Empyrean heaven, “according to what Holy
Church teacheth, who cannot lie” (Conv. ii. 3, 4). The nine moving
spheres revolve round our globe, the fixed centre of the Universe,
each of the lower eight being enclosed in the sphere above itself.
The seven lowest are the heavens of the planets: the Moon,
Mercury, Venus, the Sun, Mars, Jupiter, Saturn. The eighth or stellar
heaven, the sphere of the Fixed Stars or Firmament, is the highest
visible region of the celestial world, and to some extent corresponds
to the Earthly Paradise in the lower realms. Above this visible
firmament, the ninth or Crystalline heaven, the Primum Mobile,
directs with its movements the daily revolution of all the others. In it
nature starts; from it proceed time and motion, with all celestial
influence for the government of the world (Par. xxvii. 106-120). It is
“the royal mantle of all the volumes of the world, which is most
fervent and most living in God’s breath, and in His ways” (Par. xxiii.
112-114); and it communicates in different degrees some
participation in this quickening breath of God to the other sphere
which it encloses, and to all the Universe. It moves swiftest of all,
from the fervent desire of all its parts to be united to the Empyrean,
the spaceless and motionless ocean of Divine love, where God
beatifies the saints and Angels in the vision of His Essence. This
Empyrean is the true intellectual Paradise, for which the lower
heavens are merely sensible preparations. “This is the sovereign
edifice of the world, in which all the world is included, and outside of
which is nothing; and it is not in space, but was formed only in the
First Mind” (Conv. ii. 4); “The heaven that is pure light; light
intellectual full of love, love of true good full of joy, joy that
transcendeth every sweetness” (Par. xxx. 39-42).
 Gradations.—Each of the nine lower spheres represents a step
higher in knowledge, in love, in blessedness, until in the true
Paradise the soul attains to perfect knowledge, supreme love, and
infinite blessedness in union with the First Cause, in the Beatific
Vision of the Divine Essence. The ascent is marked by the increased
loveliness of Beatrice, as she guides Dante upwards from heaven to
heaven; it is marked, too, by gradations in the brilliancy of the
blessed spirits themselves, by their ever increasing ardour of charity
towards the poet, and by the growing spirituality of the matters
discussed in each sphere—veil after veil being drawn aside from the
mysteries of the Divine treasure-house.
The Saints.—“To show forth the glory of beatitude in those souls,”
says the letter to Can Grande, “from them, as from those who see all
truth, many things will be sought which have great utility and delight”
(Epist. x. 33). All the saints without exception have their home and
glorious seats with Mary and the Angels in that Empyrean heaven,
where they are finally seen as glorified spirit likenesses of what they
were on earth. But into each preparatory sphere, excepting the ninth,
these citizens of eternal life descend to meet Dante as, with
Beatrice, he approaches the gates of the celestial city—like the
noble soul returning home to God in the fourth and last part of life:
“And even as its citizens come forth to meet him who returns from
a long journey, before he enters the gate of the city, so to the noble
soul come forth, as is fitting, those citizens of eternal life. And thus
they do because of her good works and contemplations; for, being
now rendered to God and abstracted from worldly things and
thoughts, she seems to see those whom she believes to be with
God” (Conv. iv. 28).
In all these spheres, excepting the first, and to some extent the
second, the spirits of the blessed appear clothed in dazzling light,
which hides their proper semblances from Dante’s gaze, making
them appear as brilliant stars or flaming splendours. In the tenth
Heaven of Heavens he is supernaturally illumined, and enabled
thereby to behold them in their glorified spirit forms “with
countenance unveiled” (Par. xxii. 60, xxx. 96, xxxi. 49).
 In the three lower heavens, to which earth’s shadow was
supposed to extend (Par. ix. 118, 119), appear the souls whose lives
were marred by inconstancy in their vows, who were moved by vain
glory, or yielded to sensual love. They descend into these lower
spheres to give Dante a sensible sign of the lesser degree of the
perfection of their beatitude in the Empyrean. Domus est una, sed
diversitas est ibi mansionum; “The house is one, but there is a
diversity of mansions there.” There are different mansions of
beatitude in God’s house, proceeding from inequality in the soul’s
capacity of the Divine Charity; but in that house all are fulfilled with
the Vision of the Divine Essence, and each perfectly beatified
according to his own capacity of love and knowledge. In the spheres
of the four higher planets appear the souls of great teachers and
doctors, of Jewish warriors and Christian knights, of just rulers, of
ascetic monks and hermits; they appear as types of lives perfected
in action or in contemplation, as a sign of the different ways in which
perfection may be reached on earth and beatitude attained in
Paradise. These successive manifestations in the seven spheres of
the planets obviate what might otherwise have proved the monotony
of a single heaven, and suggest that, although each soul partakes
supremely according to its individual capacity of the Beatific Vision,
which is essentially one and the same in all, yet there are not only
grades but subtle differences in the possession of it, in which the life
on earth was a factor. In the eighth, the Stellar Heaven, still under
sensible figures and allegorical veils, Dante sees “the host of the
triumph of Christ, and all the fruit gathered by the circling of these
spheres” (Par. xxiii. 19-21), representing the Church in which these
various modes and degrees of life are brought into unison. In the
ninth, the Crystalline, the angelic hierarchies are manifested with
imagery symbolical of their office towards God and man,
representing the principle of Divine Order, the overruling and
disposition of Divine Providence in which the celestial intelligences
are the agents and instruments. The Empyrean Heaven depicts the
soul in patria, with all the capacities of love and knowledge
actualised in the fruition of the Ultimate Reality, the supreme and
universal truth which is the object of the understanding, the supreme
and universal good which is the object of the will.
 The Angels.—Each of the nine moving spheres is assigned to the
care of one of the nine angelic orders: Angels, Archangels,
Principalities; Powers, Virtues, Dominations; Thrones, Cherubim,
Seraphim. And the character of the blessed spirits that appear to
Dante in each heaven, and the subjects discussed, seem in almost
every case to correspond more or less closely with the functions
assigned by mystical theologians, especially Dionysius, St. Gregory
and St. Bernard, to the special angelic order which presides over the
sphere in question. There are two fundamental principles in the life
of the soul: nature and grace. The one is represented in the Paradiso
by the astronomical order of the heavens and their influence upon
individual disposition, furnishing man with a natural aptitude for the
moral and intellectual virtues; the other by the bounty of Divine
Grace, which reveals itself in the perfecting of the natural and the
infusion of the supernatural virtues, whereby souls become
assimilated to the angelic orders.[37] It is through these Angels (the
name is applied generally to all, as well as to the lowest order) that
God disposes the visible world; in the hands of the celestial
intelligences the heavens are as hammers, to stamp the Divine ideas
upon material creation and carry out the Divine plan in the
government of the Universe (cf. Par. ii. 127-129). And, by means of
the influence of the stars, these Angels have impressed certain men
with their own characteristics; perhaps to fill up the vacant places in
their ranks left by the fall of Lucifer’s followers, certainly to co-
operate on earth in their work. Dante himself was born beneath the
constellation of the Gemini, the glorious stars impregnated with the
virtue of the Cherubim who rule the eighth sphere (Par. xxii. 112-
123). The Cherubim represent the Divine Wisdom; their name
signifies plenitude of knowledge. According to St. Bernard, they
“draw from the very fountain of wisdom, the mouth of the Most High,
and pour out the streams of knowledge upon all His citizens.” Their
special prerogatives are fullness of Divine light, and contemplation of
the beauty of the Divine order of things; they see most into the
profound mysteries of the hidden things of God, and spread the
knowledge of Him upon all beneath them. By their inspiration Dante
co-operated in this cherubical work by writing the Divina Commedia.
The Seraphim especially represent the Divine Love. No soul appears
in the ninth heaven which they guide and in which the angelic
hierarchies are manifested; Beatrice is the sole interpreter between
the poet and the Angels, as she had been the revealer to him on
earth of Love’s “possible divinities and celestial prophecies.”
Time in Paradise.—The action of the Paradiso begins at noon,
immediately after Dante’s return from Eunoë; that is, noon on
Wednesday in Easter week in the Earthly Paradise and (the
following) midnight at Jerusalem (Par. i. 37-45). The time-references
in this third Cantica are rather doubtful (Par. xxii. 151-153, xxvii. 77-
87), but it seems probable that Dante takes twenty-four hours to
ascend through the nine material heavens to the Empyrean, which is
beyond time and space, where “the natural law in nought is relevant”
(Par. xxx. 123). When Dante woke from his “mighty trance” to the
“sound of the importunate earth,” it was perhaps about dawn on the
morning of Friday in Easter week in our world, thus completing the
seven days of his ecstatic pilgrimage, which had begun at about the
same hour on Good Friday.
Canto I.—In a lyrical prologue of stately music (Par. i. 1-36), the
poet sings of the glory of the First Mover, and prays for light and
inspiration to complete this third most arduous portion of his divine
poem. Then, in the noblest season of the year and noblest hour of
the day, as Beatrice gazes upon the sun and Dante upon her, his
mind becomes godlike, and he ascends to Heaven swifter than
lightning. To explain his ascent, Beatrice discourses upon the form
and order of God’s visible image, the Universe; and on His Eternal
Law, the sovereign plan of government existing in the Divine Mind, to
which all movements and actions of nature are subject (ibid. 103-
141). To all created things God has given an instinct, or principle of
inclination, by which, in different ways according to their nature, He
draws them all back to Himself over the great sea of being. Rational
beings alone can resist the order of the Universe and defeat the
Eternal Law by sin, which is expiated by temporary or eternal
suffering, as Dante has seen in the lower realms; but the purified
soul, in accordance with this order and law, inevitably mounts up to
find its rest in union with the First Cause. It is the doctrine of spiritual
gravitation (derived from St. Augustine), according to which the soul
is moved by love as bodies are by their weight, and all things find
their rest in order.
The Heaven of the Moon.—They are received into the eternal
pearl of the Moon (Par. ii.); where Beatrice first confutes Dante’s
former theory concerning the luminous substance of the celestial
bodies, and, by explaining how everything in the visible world
depends upon the angelic movers of the sphere, gives a mystical
interpretation of a natural phenomenon, on this first step of his
ascent to the suprasensible. Within this eternal pearl appear faint but
divinely beautiful forms of women; the souls of those who had
yielded to violence and broken their solemn vow (Par. iii.). Piccarda
Donati, sister of Corso and Forese, sets forth the perfection of
celestial charity, where all wills are made absolutely one with the will
of God, who has awarded different degrees or mansions of beatitude
to all His chosen ones:

E la sua volontade è nostra pace,

“And His will is our peace.”[38] Transfigured now with ineffable joy,
Piccarda tells the pathetic story of her frustrated life on earth; and
points out to Dante the Empress Constance, mother of Frederick II.,
torn, like her, from the convent’s shelter. Beatrice explains to the
poet the place of all the saints in the Empyrean—the “heaven of
humility where Mary is,” as Dante had sung long before of Beatrice
herself in the Vita Nuova—and the reason of this temporary
apparition in the moon (Par. iv.). The other questions solved in this
sphere are all connected with Free Will. Rectitude of will is
necessary for the gaining of Paradise, and nothing whatever can
take away that freedom of the will. “As regards the proper act of the
will, no violence can be done to the will”; and, since Piccarda and
Constance yielded through fear of greater evil, they fell voluntarily
from the state of perfection to which they were called. Freedom of
the will is God’s greatest gift to man (Par. v. 19-24); hence the
sanctity of an accepted vow, wherein this supreme gift is offered to
God as victim, although Holy Church has power to commute, save,