392
2, FEBRUARY 2011
A Physics-Based Three-Dimensional Analytical
Model for RDF-Induced Threshold
Voltage Variations
Georgios Panagopoulos, Student Member, IEEE, and Kaushik Roy, Fellow, IEEE
Abstract—In this paper, a 3-D analytical model is proposed to
capture the threshold voltage, surface potential, and electric field
variations induced by random dopant fluctuations in the channel
region of metal–oxide–semiconductor field-effect transistors. The
3-D model treats the effect of each dopant separately and is
based on fundamental laws of physics. The proposed approach
enables determination of transistor threshold voltage variations
with both very low computational cost and high accuracy. Using
the developed model, we performed statistical analysis, simulating
more than 100 000 transistor samples. Interestingly, the results
showed that, although the distribution of the threshold voltage
for large-channel transistors is Gaussian, for scaled transistors,
it is non-Gaussian. Furthermore, the proposed model predicts
known formulas, which are proven for 1-D analysis and large
transistors, simply by setting the appropriate transistor size. As
a consequence, this model is a logical extension of the theory of
large transistors to nanoscaled devices.
Index Terms—Electric field, random dopant placement (RDF),
surface potential, Vth variations, 3-D analytical model.
I. I NTRODUCTION
A GGRESSIVE transistor scaling has led to random varia-
tions in the transistor threshold voltage [1]–[8], [17]–[23].
This is mainly due to random phenomena such as random
dopant fluctuations (RDFs) in the device channel region (fluc-
tuations in both the number and location of dopant atoms)
[1]–[3], [17]–[23], line edge roughness [4], [5], and work func-
tion variations [6]. It has been shown that RDF renders the most
in the randomness of transistor characteristics [1]–[3]. It should
be noted that RDF modeling not only requires knowledge of the
number of dopants but also information about the location of
the individual dopant atoms in the metal–oxide–semiconductor
field-effect transistor (MOSFET)’s channel region [1]–[3].
Hence, development of 3-D models is necessary to capture the
effects of RDF. This randomness in the profile of dopant atoms
affects both the electrostatic potential and the electric field
at the Si-SiO2 interface of the transistor. Consequently, RDF
affects the threshold voltage. In turn, this randomness varies
Manuscript received June 30, 2010; revised October 24, 2010; accepted
November 3, 2010. Date of publication December 13, 2010; date of current
version January 21, 2011. The review of this paper was arranged by Editor
J. C. S. Woo.
The authors are with the School of Electrical Engineering, Purdue Uni-
versity, West Lafayette, IN 47907 USA (e-mail: gpanagop@ecn.purdue.edu;
kaushik@ecn.purdue.edu).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TED.2010.2093140
0018-9383/$26.00 © 2010 IEEE
IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 58, NO.
circuit parameters that are important for the design of digi-
tal and analog circuits [25]–[28]. For instance, the transistor
threshold voltage Vth affects the delay and the on/off cur-
rent of a digital circuit, and the vertical electric field affects
the mobility of the carriers inside the channel and, thus, the
transconductance.
Variations of the threshold voltage have been ruminated upon
not only experimentally [24] but also numerically through the
use of 3-D atomistic [1]–[3] or quantum-level simulators [30],
[31]. The main disadvantage of these models is that they require
large computational effort and/or use complex techniques for
threshold voltage determination [12]–[16]. On the other hand,
there exist analytical models [9], [10], but these models are not
very accurate, because they fail to predict the non-Gaussian
Vth distribution for scaled technologies. Hence, average and
standard deviation is not enough statistical information to de-
scribe the statistics of Vth . Nonzero skewness and kurtosis
throw more light on the real Vth distribution, and our proposed
model provides this information. In addition, in this framework,
we not only propose a model for Vth but also provide statistical
information for other important quantities such as the electric
field and electrostatic potential.
In this paper, the variations of many important transistor
characteristics such as the threshold voltage, the electric field,
and the effective doping concentration are modeled using the
fundamental laws of physics. The salient features of the pro-
posed model are given as follows: 1) It is 3-D, i.e., it takes
into account the exact location of the dopants in the 3-D space.
2) It is quasi-atomistic, i.e., it treats the charge of each dopant
separately. 3) It is simple; hence, it gives the intuition for the
origin of the variations and their effects on the transistor char-
acteristics. 4) It is analytical, i.e., expressions in closed form are
developed. Finally, a framework is established where quantities
such as the total number of dopants, the electrostatic potential,
and the electric field are defined based on this framework.
The preceding features allow fast and accurate Monte Carlo
simulations of RDF in nanoscaled transistors.
This paper is organized as follows. In Section II, the proposed
framework is presented, demonstrating the physical laws on
which it is based on. The appropriate assumptions under which
the model is derived are also stated in this section. The model
is completed with the derivation of modified expressions for
important quantities: electric field, electrostatic potential, and
threshold voltage under the effect of RDF. In the next section,
we convert the solutions to discrete form, and in Section IV,
PANAGOPOULOS AND ROY: ANALYTICAL MODEL FOR RDF-INDUCED THRESHOLD VOLTAGE VARIATIONS
Fig. 1. Flowchart illustrating the methodology that is used for modeling RDF.
the statistical analysis of the threshold voltage using the pro-
posed framework is presented. In Section V, the results of
our methodology are demonstrated, and a comparison between
the proposed model and other atomistic-level simulators and
experimental results are shown. Finally, the conclusions are
drawn in Section VI.
II. RDF: T HE F RAMEWORK
In this section, let us establish the framework for modeling
the effect of RDF on nanoscaled MOSFETs. This framework
is illustrated using the flowchart in Fig. 1. In the following
paragraphs, we describe the individual steps for modeling the
threshold voltage variations considering RDF in detail. These
steps are denoted by the letters A–E in Fig. 1.
A. Input Parameters
The width W and the length L of a transistor constitute the
size-related input parameters. The width is obtained from the
circuit-level design, and the length is specified for a technology
node. The third input parameter is the nominal doping concen-
tration NA or, in general, the channel doping profile. Finally,
the oxide thickness Tox completes the list of input parameters
needed for deriving the proposed model. These input parame-
ters are sufficient to determine the characteristic behavior of a
transistor and grasp the trends of a specific technology node, as
it is shown later in this paper.
B. Depletion Width
In the next step of our framework, we determine the max-
imum depletion width Wdm of a transistor. For this determi-
nation, it is essential to know not only the number but also
the position of the ionized and neutral dopants in the channel
region since the depletion width is expected to vary spatially
due to nonuniformity of the doping profile. At this point, we
need to make two observations: 1) The contribution of the
393
ionized dopants away from the surface and close to the edge of
the depletion region is not as significant as compared to those
close to the oxide interface. 2) On the average, the dopants
that are ionized and outside the depletion region (as previously
defined) are canceled out by those that are neutral and inside
the depletion region. These two observations directly lead to the
approximation of the maximum depletion width, as expressed
by the following equation, that is shown for 1-D uniform
channel region [32]–[34]:

Wdm = 4εsi kT ln(NA /ni )/q 2 NA (1)
where q is the electronic charge, εsi is the permittivity of silicon,
k is the Boltzmann constant, T is the temperature, NA is the
doping concentration, and ni is the intrinsic carrier density.
Consequently, (1) is an excellent approximation, even in the
case of scaled transistors. This assumption is essential to elim-
inate the otherwise needed iterations for the exact calculation
of the spatially varying depletion width. As we discuss in the
next section, the depletion region is modulated by the drain
voltage.
C. Channel Region
In the third step of the proposed framework, the transistor
size (W and L), along with the maximum depletion width Wdm
given by (1), form the channel region, as illustrated in Fig. 2(a)
and (d). This region is meshed, and at each vertex of this mesh,
the following occur: 1) An identification number i is given.
2) A binary random variable Xi is assigned to denote the
existence or absence of a dopant, with 1 < i < N , where N
is the total number of possible positions that a dopant can be
placed in. 3) The 3-D coordinates of each dopant in the R3
space are determined. The assignment of the value “1”/“0”
to the random variable Xi denotes the existence/absence of
a dopant at the corresponding vertex given by the pertinent
coordinates. For the sake of simplicity and demonstration of the
method, let us assume that all the random variables Xi follow
the Bernoulli distribution (coin toss) [35], i.e.,

1, dopant existence
Xi = . (2)
0, dopant extinction
For a uniform doping profile, the associated probabilities
can be defined as P [Xi = 1] = s, P [Xi = 0] = 1 − s = t, and
their corresponding first and second statistical moments can
be obtained as E[Xi = 1] = s and Var[Xi ] = s(1 − s) = s · t,
respectively [35]. As will be explained in Section IV, the
probability s depends on the nominal doping concentration of
the channel region NA .
It is known that, for scaled devices, nonuniform doping
profile is necessary in order to both control the threshold
voltage and reduce the short-channel effects. In literature, many
nonuniform doping profiles have been proposed (mainly us-
ing ion implantation techniques) to overcome these problems.
Our framework incorporates nonuniform doping profiles by
appropriately assigning the probabilities si and ti for each
random variable Xi . These probabilities correspond to the
394 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 58, NO. 2, FEBRUARY 2011

Fig. 2. Two random patterns from the channel region of transistors of the same technology node (22 nm) showing the random number and position of dopants.
The corresponding 3-D potential and electric field at the surface are illustrated in sequence.

process/doping method and the doping profile used (i.e., uni- in the next two sections, which are necessary to build our
form, step, retrograde, and Gaussian) [34]. For example, if a model. These variables are categorized as quantitative variables
Gaussian doping profile is assumed, then the concentration and (number of dopants and effective doping concentration) and
the corresponding probabilities si are given by the equations electrical variables (surface potential, electric field at the sur-
face, and threshold voltage). All these variables are represented
N0 − (z−z2c )2 a3 N0 − (zi −z2c )2
NA = √ e 2σ si = √ e 2σ as “leaves” in the flowchart (Fig. 1) to designate that they are
2πσ 2πσ the outputs of the model.
where zc is the center of the Gaussian doping profile, σ is its
deviation from the center, and a is the lattice constant (the min- D. Modeling Quantitative Variables
imum possible distance between two dopants). In Section IV,
it becomes more obvious why we have to multiply NA (x, y, z) The quantitative variables give intuitive information about
by the term a3 to determine si . the number of dopants in the channel region, which is an
Note that the maximum depletion width Wdm given by important and critical quantity for nanoscaled transistors.
(1) is not valid for nonuniform doping profile, but generally, 1) Number of Dopants Dp : The first random variable Dp
there are analytical expressions for Wdm for some common denotes the “total number of ionized dopants in the channel
profiles. Furthermore, for doping profiles where it is difficult region” and is defined as the sum of Xi over all the N mesh
to analytically determine Wdm , a technology computer-aided vertices
design device simulator (Medici, Taurus, Sentaurus) can be
used to extract Wdm before the analysis starts. 
N
Dp = Xi . (3)
In order to understand the concept of our framework, let
i=1
us consider an example [Fig. 2(a) and (d)] of two differ-
ent transistor instances at the same technology node (W = 2) Effective Doping Concentration Cp : The “effective dop-
L = 22 nm), and let us assume uniform doping concentra- ing concentration” is defined as the ratio of the total number
tion (NA = 1019 cm−3 ). These channel region instances are of ionized dopants in the channel region to the volume of this
generated by performing two iterations, as shown in Fig. 1. region. It is evident that the effective doping concentration of
In addition, we define a vector {X} of the aforementioned each transistor is also a random variable that is related to the
random variables Xi : {X} = { X1 X2 · · · XN }T . In the predefined random variable Dp by
subsequent sections, this vector will be very useful for the
formulation of the problem.
1 
N
Dp
Having discussed the tools that are required to build a 3-D Cp = = Xi . (4)
analytical model for RDF, various random variables are defined W · L · Wdm N a3 i=1
PANAGOPOULOS AND ROY: ANALYTICAL MODEL FOR RDF-INDUCED THRESHOLD VOLTAGE VARIATIONS 395

The effective doping concentration Cp is different from the

nominal doping concentration NA due to random number of
ionized dopants. As it is shown in Section IV, the average value
of Cp is equal to NA . The reason that both Cp and Dp are
defined becomes evident later in Section IV, where we discuss
about their statistics.

E. Modeling Electrical Variables

Before defining the electrical variables, it is implicit to
mention the laws of physics that the proposed model is based
on. The use of Maxwell equations and, more specifically, the
Gauss law or the Poisson equation is not helpful enough to
provide analytical expressions for the electrical variables due
to the lack of symmetry in this problem. Rather, the use of
alternative forms of Gauss equation is more appropriate and
convenient in this case. Thus, in this paper, the electric field
and the potential are determined by Coulomb’s law [36], [37]
Qi Qi
E= Φ= (5)
4πεSi Ri2 4πεSi Ri
where Qi is the charge of the ith dopant, and Ri is the distance
of the ith dopant to the point where the electric field and
potential are to be calculated.
By applying the preceding two fundamental relations, ana-
lytical expressions for the electrical properties of the transistor
are derived [(6)–(10)]. In addition, an approximation commonly
referred to as “delta depletion approximation” [32]–[34] is also
utilized in our model, which states that there are no mobile
carriers/charges in the channel region for gate voltages less than
the threshold voltage.
The most important electrical quantities of a transistor are
its electrostatic potential and the electric field at the Si-SiO2
interface (Fig. 3) when the applied gate voltage is equal to the
threshold voltage. Both of them are functions of the coordinates
at the points that they are calculated. Their expressions are
essential for the definition of threshold voltage. Note that,
due to the random position of the dopants, both the electric
field and the electrostatic potential at the Si-SiO2 interface
are spatially nonuniform. These quantities are also expected
to be spatially correlated, i.e., they are dependent on their
corresponding values at the neighboring points (see Fig. 2). Let
us now define the electrical variables beginning with the surface
potential.
1) Surface Potential Φ: The surface potential is determined
using the superposition principle and is given as [34]
Ψ(x, y) = Φ(x, y) + U = Φ(x, y) + US + UD + UB . (6)
Φ is the contribution to the potential due to random dopants, and
US , UD , UB are solutions that satisfy the boundary conditions
at the source, drain, and body electrodes, respectively.
The surface potential Φ at a point (x, y) at the Si-SiO2
interface is given by the sum of the potential contribution of
each ionized dopant in the channel. This can be expressed as

N region as VDS increases (see Fig. 3).
Φ(x, y) = (Φi (x, y) + Φi (x, y)) (7)
i=1
Fig. 3. Vertical cross section illustrating real and image random dopants used
for modeling RDF. The depltion width is modulated by Vds .
where Φi (x, y) is the contribution of the ith dopant (if it is
ionized) to the surface potential at position (x, y), and Φi (x, y)
is the contribution of the corresponding image (or mirror)
charge, which results from the method of images [36], [37]
−qXi qXi
Φi (x, y) = Φi (x, y) = . (8)
4πεSi Ri (x, y) 4πεSi Ri (x, y)
In (8), Ri is the distance from a dopant i to a point (x, y)
on the surface, and Ri is the corresponding distance of the
image charge (see Fig. 3). Note that (8) are valid for n-channel
metal–oxide–semiconductor (MOS) transistors with p-type
substrate. For p-channel MOS transistors, (8) are the same but
with opposite signs. In the preceding formulas, xi , yi , and zi
are the coordinates of the point where the dopant with id i is
located. Φi in (7) ensures that both the potential and the electric
field at the end of the depletion region approach zero. The
application of this method is illustrated in Fig. 3. The transistor
in this figure has seven random dopants (D1 − D7 , A1 − A6 ).
Note that the potential and the electric field at a point away from
the depletion width boundary and closer to the body electrode
tends to zero. This is evident from (7) as the contribution
of a random dopant to the potential (Φi ) is canceled out by
the contribution of a symmetrical image dopant (Φi ). This
cancellation occurs from the placement of the image random
dopants (I1 − I7 ) at symmetrical positions (Fig. 3). In addition,
it is also clear from (7) that the potential is a function of both
the number and the position of the random dopants. Note that
additional dopants (A7 − A9 ) are included in the depletion
Having discussed about Φ, the rest of the terms of (6) are
given by solving the Poisson equation with voltages at the
396 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 58, NO. 2, FEBRUARY 2011

contacts as boundary conditions

   
π(L−x) πL
US =b1 sinh / sinh
Wdm +3Tox Wdm +3Tox
   
π(L−x) πL
UD =c1 sinh / sinh
Wdm +3Tox Wdm +3Tox

b1 =4/πVbi − 2/π(1 − 4/π 2 )(1+6Tox /Wdm )Φ(x, y)

4 2
c1 = (Vbi +Vds )− (1−4/π 2 )(1+6Tox /Wdm )Φ(x, y).
π π
UB is an order of magnitude smaller than US and UD and has
been ignored in this analysis. The complete derivation is given
Fig. 4. Two-dimensional discrete lattice at Si-SiO2 interface used for the
in [34, Appendix A]. conversion of the electrical variables to matrix form.
2) Electric Field EZ : The electric field is a vector and
consists of three components (x, y, and z directions). Even maximum surface potential across the “lane.” Such formulation
though all these components can be calculated from the pro- is described mathematically as
posed model, the most important for the determination of the
threshold voltage is the component vertical to the Si-SiO2 Yi (y) = max (Vt (x, y)) (12)
x=(0...L)
interface (z-component). Hence, the electric field is given by
the following equation: for all y along the channel width. Finally, since we have
assumed that all the “lanes” have the same width dy, the

N
  transistor’s threshold voltage is defined as the average of all the

EZ (x, y) = EZ,i (x, y) − EZ,i (x, y) (9) lane threshold voltages [23]
i=1
W W  

where EZ,i (x, y) and EZ,i (x, y) are defined as 1 1 εsi EZ
Vth = Yi (y)dy = max Ψ+ dy. (13)
W W x=0...L Cox
qzi Xi  qzi Xi 0 0
EZ,i (x, y) = EZ,i (x, y) = . (10)
4πεSi Ri3 4πεSi Ri3 Equation (12) is the analytical expression for the threshold
voltage. Alternatively, the threshold voltage can be defined by
3) Threshold Voltage Vth : In the final step of the proposed averaging all over the channel. In this case, the definition of Vth
framework, we define the threshold voltage Vth . The definition is given by
of the threshold voltage is not unique; it can be extracted from
current–voltage characteristics [8], [9], or it can be computed W L  
1 εsi EZ (x, y)
using percolation paths, which connect the source and the drain Vth,lin = Ψ(x, y) + dy. (14)
terminals [5], [8]. WL Cox
0 0
Before discussing the modeling of Vth , we have to define
the random variable Vt (x, y), which represents the gate voltage This definition can be used when Vds is small (i.e., when
that should be applied in order to invert the channel locally. linear threshold voltage Vth,lin is of interest). On the other
Since the electrostatic potential Ψ and the electric field EZ at hand, the proposed framework does not limit us using other
the surface are known [(6) and (9)], the local Vt can be defined definitions that potentially improve the accuracy. Equation (12)
as the voltage drop across the bulk plus the voltage drop across is more general than (14) but more complex for statistical
the oxide analysis. Note that, for large transistors, (12) and (14) become
equivalent; thus, either of them can be used. Note that all the
εsi EZ (x, y) equations defined in this section are functions of the input
Vt (x, y) = Ψ + Vox = Φ(x, y) + U + (11)
Cox parameters W , L, NA , and Tox . Now, we can present the
random variables previously defined in discrete form.
where Cox = ox /Tox . In this paper, we simplify the definition
of threshold voltage by assuming that the electrons move from
drain to source contact laterally. This assumption reduces fur- III. S OLUTIONS IN D ISCRETE F ORM
ther the complexity and has minor impact on the accuracy of In this section, we discretize the electrical variables in order
the model, as can be observed from the obtained results shown to perform easier statistical analysis and obtain numerical solu-
later. Such simplification is allowed by dividing the channel in tions. In order to discretize Φ and EZ , we use a 2-D discrete
parallel “lanes” and assuming that electrons do not change their lattice similar to the one shown in Fig. 4. In this case, the
“lane” during their transport from source to drain contacts [23]. discrete form of Φ and EZ can be written in matrix form as
This is equivalent to a 1-D channel (Landauer model), in which
the threshold voltage is easily determined by just taking the {Φ} = [A] · {X} {EZ } = [B] · {X} (15)
PANAGOPOULOS AND ROY: ANALYTICAL MODEL FOR RDF-INDUCED THRESHOLD VOLTAGE VARIATIONS 397

where Φ(i,j) and EZ(i,j) are the potential and electric fields on depends on the input parameters W , L, and NA . For instance,
the Si-SiO2 interface around the mesh node (i, j), respectively. an increase in N translates to an increase in either the transistor
Note that K and L are the numbers of discrete lattice points size (W and/or L) or the depletion width (Wdm ), which, in
in the x and y directions, respectively (Fig. 4). In addition, turn, depends on input NA . Since the random variables Xi
this representation becomes exact (with the continuous case are independent, the distribution of Dp follows the Binomial
previously described) only in the limit, where the grid distances distribution, which approaches the Poisson distribution as N
a and b approach zero. However, as long as a and b are increases. Moreover, Dp approaches the normal distribution
smaller than the spatial scale on which Φ and EZ varies, this (Dp ∼ Normal(μDp , σDp )) as N becomes very large, accord-
representation is reasonably accurate. The size of the matrices ing to central limit theorem (CLT) [35]. Note that μDp and σDp
[A] and [B] is (KL)xN , and their corresponding elements are increase with the enlargement of the channel region (W , L,
of the form and Wdm ), whereas SDp decreases (approaches to zero) as the
q channel region increases. This verifies the general intuition that
α(κ,λ)i = (1/Ri (xκ , yλ ) − 1/Ri (xκ , yλ )) (16) large transistors are uniform.
4πεSi
qzi  
b(κ,λ)i = 1/Ri3 (xκ , yλ ) − 1/Ri3 (xκ , yλ ) . (17)
4πεSi B. Effective Doping Concentration Cp
Using the preceding formulation for Φ and EZ , we can define Following the same arguments used for the random variable
the discrete form of Vt that will allow us to calculate the Vth of Dp , the random variable Cp also approaches the normal distri-
a transistor numerically. Substituting (15) into (11), the local bution (Cp ∼ Normal(μCp , σCp )) as N increases (CLT). The
threshold voltage {Vt } can be written as a function of the mean value and the standard deviation are given by
random vector {X} as √
  N
1 s st
εSi εSi μCp = E[Xi ] = 3 = NA σCp = √ (21)
{Vt } = {Ψ} + {EZ } = U + [A] + [B] {X}. N a3 a 3
a N.
Cox Cox i=1
(18)  √
The spread of Cp is SCp = t/N s ∝ 1/ N . With the aid of
Finally, the {Vt } vector is utilized for the determination of (21), the probability s of a Bernoulli random variable Xi is
Vth for low [(14)] and high Vds [(13)], i.e., determined and is a function of the doping concentration NA ,
i.e., s = a3 NA . In contrast to the mean of Dp , the mean of Cp
1  1
K Λ Λ
Vth,lin = Vt Vth = max Vt(i,j) . (19) is constant and is equal to NA , which is the nominal doping
KΛ i=1 j=1 Λ j=1 i=1...K concentration. We also observe that both σCp and SCp approach
to zero as N becomes very large.
Note that all the variables defined in this section depends on the
parameter matrices [A] and [B] and the oxide thickness Tox .
For a specific technology node, [A] and [B] are calculated only C. Surface Potential Φ
once, and their size depends on the size of the device under Having discussed the statistics of the quantitative variables
analysis. A more accurate representation can be easily achieved Dp and Cp , we will next present the statistical analysis of the
by making the grid finer. In the next section, we elaborate the electrical variable Φ. Since Φ is a vector of random variables,
statistical analysis of the random variables. its mean and variance are also expected to be vectors with the
elements, mean, and variance of Φ(i,j) , respectively. Using the
IV. S TATISTICAL A NALYSIS definition of Φ [(15)], we derive the mean of Φ as

T
In this section, we calculate the mean, the variance, and 
N 
N

the distributions of the random variables previously defined. In E[Φ] = s α(1,1)r · · · α(K,L)r ≡ [MΦ ]. (22)
addition, a discussion on the theoretical results is presented. r=1 r=1

Each element of the new vector MΦ is the average surface po-

A. Number of Dopants Dp tential at each grid point on the Si-SiO2 interface. Furthermore,
the linearity property of the covariance operator applied on (15)
In Section II, the random variable Dp was defined as the yields
finite sum of the independent random variables Xi (3). The
mean μDp and the standard deviation σDp of Dp are given by Cov[Φ] = s · t[A][A]T ≡ [ΣΦ ]. (23)
the following relationships:

N
√ In addition, the random variable Φ follows the normal dis-
μDp = E[Xi ] = N · s σ Dp = N · s · t. (20) tribution since it is the sum of independent Bernoulli variables
i=1 (not identical). Hence, Φ(κ,λ) ∼ Normal(μ(κ,λ) , σ(κ,λ) ), where

The spread (coefficient of variation) is SDp = σD /μD = N
 √ N 
t/N s ∝ 1/ N . From the preceding equations, we observe μΦ(κ,λ) = s a(κ,λ)i σΦ(κ,λ) = st a(κ,λ )i (24)
that μDp , σDp , and SDp are functions of parameter N , which i=1 i=1
398 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 58, NO. 2, FEBRUARY 2011

The joint probability density function (pdf) for the potential

vector using the matrices MΦ and ΣΦ is given approximately by
T −1
a normal pdf [35]: f (φ) = exp(−(1/2)(φ − MΦ ) ΣΦ (φ −
MΦ ))/ (2π) K·L |ΣΦ |. The pdf function f (φ) shows that the
potentials at neighboring grid points are correlated since ΣΦ is
not diagonal matrix.

D. Vertical Electric Field EZ

In a similar way, we derive the statistics of the second statis-
tical variable EZ , which is a vector of random variables. Hence,
its mean and variance are expected to be vectors with elements,
mean, and variance of EZ(i,j) , respectively. The mean of EZ at
each grid point applying the mean and covariance operator on
(9) is derived as

N T
 
N
E[EZ ] = b(1,1)r · · · b(K,L)r ≡ [ME ] (25)
r=1 r=1

Cov[EZ ] = B · Var[X] · B = st[BBT ] ≡ [ΣE ].

T
(26)

Furthermore, the random variable EZ follows the normal

distribution since it is a sum of independent Bernoulli variables.
Therefore, EZ(κ,λ) ∼ Normal(μE(κ,λ) , σE(κ,λ) ), where

N
N 
μE(κ,λ) = s a(κ,λ)i , σE(κ,λ) = st b(κ,λ)i . (27)
i=1 i=1

The joint pdf for the potential vector using the known constant
matrices ME and ΣE is given approximately by the multi-
variablenormal pdf [35]: f = exp(−(1/2)( − ME )T Σ−1 E ( −
ME ))/ (2π)K·L |ΣE |. Fig. 5. Statistics of the quantitative random variables. (a) Number of dopants
Note that the electric fields at neighboring grid points at the in the channel Dp . (b) Effective doping concentration Cp obtained for different
Si-SiO2 interface are correlated. This correlation is one of the values of NA and transistor size.
most important advantages of our model and is necessary for
accurate determination of the electric field and the threshold V. R ESULTS AND D ISCUSSIONS
voltage. In order to show the efficacy of the proposed model, we
performed 100 000 Monte Carlo simulations on various-sized
E. Threshold Voltage Vth transistors with different doping profiles and obtained distribu-
tions and statistical data for a number of parameters.
Finally, we derive expressions for the statistics of the thresh- In Fig. 5, the mean and the standard deviation of Dp and Cp
old voltage Vth . Using (19), the mean of the threshold voltage with different values of W and L are demonstrated. Fig. 5(a)
is derived as follows: verifies the linear relationship of μDp with W and the square
K Λ   root of σDp with W , as captured by (20). We can also observe
1  εSi
E[Vth ] = U + MΦ + ME ≡ μVth . from Fig. 5(b) that μCp is constant and is equal to NA while
√
KΛ i=1 j=1 Cox
σCp is proportional to 1/ W L, as predicted by (21). Since,
(28) for very large transistors, σCp decreases and tends to zero, we
may conclude that the effective doping concentration is given
μVth is the average value of threshold voltage and depends on
by a single value NA (as in the 1-D analysis). Furthermore, we
the matrices MΦ and ME . Similarly, applying the covariance
observe that transistors with higher doping concentration suffer
operator on (19), we obtain the standard deviation of Vth
from larger variations in the number of dopants.

Λ   The distributions of the threshold voltage for different tran-
s · t  K 
ε2 2εSi
σVth = 
sistor sizes are shown in Fig. 6. We observe that the variations
AAT + Si 2
BB T
+ AB T
.
KΛ i=1 j=1 Cox Cox of the Vth due to RDF increase as the transistor size decreases.
(29) In addition, our model predicts the skew distribution of Vth
that has been observed by the authors in [22] and [29] and is
Note that σVth depends on the matrices A and B. attributed to the small number of dopants in the channel region.
PANAGOPOULOS AND ROY: ANALYTICAL MODEL FOR RDF-INDUCED THRESHOLD VOLTAGE VARIATIONS 399

Fig. 8. Threshold voltage distributions of a sample set of 250 microscopically

different transistors obtained changing the model parameters Tox with W =
L = 50 nm and NA = 5 × 1018 cm−3 .
Fig. 6. Threshold voltage distributions obtained changing the model parame-
ters length and width (W = L = [22 nm−200 nm]).

Fig. 7. Logarithmic plot of the threshold distribution for short- and long-
channel transistors obtained from the proposed model for a sample set of
100 000 MOSFETs with Tox = 3 nm and NA = 5 × 1018 cm−3 .

This assymetry becomes more significant as the transistor size

or doping concentration decreases as it is evident in the log
scale plot shown in Fig. 7.
Tox is another important input parameter that affects Vth .
In Fig. 8, we plot the threshold voltage of a sample set of
250 microscopically different transistors for Tox = 1 nm and
Tox = 3 nm. We observe that the increase in Tox implies larger
variations of Vth . In addition, in this figure, we observe that
the number of the dopants in the channel is not constant and
that there exists threshold voltage variations even for transistors
that have the same number of dopants in their channel. This is
attributed to the fact that their position plays a significant role
for the determination of Vth .
In Figs. 9–11, we demonstrate statistical information pro-
Fig. 9. Comparison of the σVth obtained from the proposed model, the
vided by our model, together with the analytical expression analytical expression in [10], and the “atomistic” simulator in [2] for 100 000
proposed in [10] and results extracted from the “atomistic” MOSFETs with different (a) L < 100 nm and (b) L < 400 nm.
400
Fig. 10. Comparison of the σVth obtained from the proposed model, the
analytical expression proposed in [10], and the “atomistic” simulations in [1]
for a sample set of 100 000 MOSFETs with Tox = 3 nm and W = L = 50 nm
and for different values of NA .
Fig. 11. Comparison of σVth obtained from the proposed model, the analyt-
ical expression proposed in [10], and the “atomistic” simulations in [1] for a
sample set of 100 000 MOSFETs with different Tox , keeping the size constant
W = L = 50 nm and NA = 5 × 1018 cm−3 .
simulator proposed in [2]. From this comparison, it is evident
that our model is accurate since it closely follows the results
obtained from the “atomistic” level simulations. In addition, the
Vth variations that are predicted by our model are verified by
chip measurements [24]. Note that the comparison with other
models and data from chip measurements is not absolutely
correct since we do not know the exact structure and doping
profiles of these transistors, even though we observe that our
model follows not only the trends such as σVth but also skew-
ness and kurtosis.
The design of large circuits (i.e., static random-access mem-
ories) requires knowledge of Vth distribution at the tails. Hence,
standard deviation is not sufficient to determine the exact shape
of this distribution. Rather higher order statistical moments are
needed. These quantities are the skew and kurtosis and are
appropriate to describe the assymetry and the shape at the tails
of a distribution. Note that, for accurate skew and kurtosis, the
sample size should be large (more that 100 000) in order to
achieve very small statistical error. Moreover, the results are ob-
IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 58, NO. 2, FEBRUARY 2011
tained in reasonable time (few hours), as compared to the time
needed by the “atomistic simulator” (approximately 20 cpu
years [22]). In Fig. 13, we plot the values of skewness and
kurtosis of Vth that are predicted by our model for different
transistor sizes. We observe that, for transistors with L <
45 nm, there is a significant increase in skew to the left (positive
skew), and for transistors with L < 20 nm, there is a dramatical
increase in kurtosis. As we have already explained, Vth also
depends on the external operating conditions, i.e., the voltage
difference between drain and source Vds . The application of
high Vds has two effects on Vth distribution. The first one is
the reduction of the mean of Vth and affects all the nanoscaled
transistors. This reduction is attributed to the well-known effect
(from 2-D analysis) that is called Vth rolloff. According to that,
the mean of Vth distribution decreases.
The second effect is relevant to RDF and affects the other
statistical moments, like standard deviation, skewness, and kur-
tosis. In this paper, for the first time, we show that drain-induced
barrier lowering (DIBL) increases all of these moments, and
this is attributed to the fact that the drain potential nulifies the
effect of dopants that are located close to the drain contact.
It is evident that, if there are many dopants close to the drain
side, then the potential profile close to drain will be high. This
increased potential is interpreted as high Vth if Vds is low. If
Vds is high, then the shape of the potential profile close to drain
does not play a significant role in Vth determination since the
potential in the drain region has decreased and does not obstruct
the motion of electrons. For instance, the transistor with low
Vds , as shown in Fig. 12(a) and (b), has smaller Vth than the
same transistor operating at high Vds due to the significant drop
of the potential close to drain. This phenomenon is not observed
for the second sample [Fig. 12(c) and (d)], because the potential
profile drops close to source contact. Thus, we conclude that
DIBL increases the variability for the nanoscaled transistors.
This increase in variability can be explained intuitively by
observing that the part of the channel that is controlled by the
gate (that is close to source) becomes smaller for high Vds .
Hence, a smaller channel is affected more by RDF variations.
In order to compare the effect of Vds on Vth distributions, we
have plotted the standard deviation, skewness, and kurtosis for
low and high Vds in Fig. 13. In addition, the Vth distributions
for 22-nm transistors operating at low and high Vds are depicted
in Fig. 14.
VI. C ONCLUSION
In this paper, we have presented a physics-based 3-D ana-
lytical model for threshold voltage variations considering RDF.
The model provides analytical expressions and statistical analy-
sis for electrical quantities (surface potential, electric field,
and threshold voltage) with respect to the number/location of
dopants and effective doping concentration. The validity of
these expressions is verified using samples of 100 000 micro-
scopically different MOSFETs. The effects of L, W , NA , and
Tox on the random dopant-induced threshold voltage variations
are studied. The results of the proposed framework are com-
pared with a 3-D atomistic simulator, another analytical model,
and chip measurements. Although the analytical model predicts
PANAGOPOULOS AND ROY: ANALYTICAL MODEL FOR RDF-INDUCED THRESHOLD VOLTAGE VARIATIONS 401

Fig. 12. Three-dimensional band diagram for low and high Vth of two 20-nm transistors is illustrated in sequence, demonstrating their correlation.

Fig. 13. Comparison of the Vth for low and high Vds obtained from the proposed model for a sample set of 100 000 MOSFETs. (a) Standard deviation. (b)
Skewness. (c) Kurtosis.

the general trends of threshold fluctuations due to RDF, it fails case (1-D analysis). This convergence is achieved by increasing
to predict quantitatively the magnitude of these fluctuations and the transistor size to infinity. For mathematical simplicity and
higher order statistical moments. On the other hand, the 3-D without loss of generality, we suppose that the shape of the
“atomistic simulator” predicts quantitatively the magnitude of transistor channel is cylindrical (Fig. 15).
the fluctuations at the cost of extended simulation time. Starting from (7) and applying the average operator on it, we
Finally, the proposed model succeeds in capturing the 3-D obtain
effects, and it can be used for predicting electrical properties
NA 
N
of devices in the nanometer and micrometer regime. It is a
E[Φi ] = air
complete and unified model that is able to capture the effect of a3 r=1
RDF for all the technology nodes. This is achieved mainly by ⎡W
effectively calculating the electrostatic potential, electric field, ρ 2π dm
∼ qNA ⎣ rdrdϕdz
and threshold voltage. = 
4πεSi r2 + (z + b)2
0 0 0
⎤
A PPENDIX A Tox
rdrdϕdz ⎦.
In this section, we show that the variables Φ and EZ converge − 
r2 + (z + b)2
to a value that is the same as that we know for the uniform −Wdm −Tox
402 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 58, NO. 2, FEBRUARY 2011

[4] S. Xiong and J. Bokor, “A simulation study of gate line edge

roughness effects on doping profiles of short-channel MOSFET de-
vices,” IEEE Trans. Electron Devices, vol. 51, no. 2, pp. 228–232,
Feb. 2004.
[5] Y. Ye, F. Liu, S. Nassif, and Y. Cao, “Statistical modeling and simulation
of threshold variation under dopant fluctuations and line edge roughness,”
in Proc. DAC, 2008, pp. 900–905.
[6] C. Hwang, T. Li, M. Han, K. Lee, H. Cheng, and Y. Li, “Statistical
analysis of metal gate workfunction variability, process variation, and
random dopant fluctuation in nano-CMOS circuits,” in Proc. SISPAD,
2009, pp. 1–4.
[7] B. Hoeneisen and C. A. Mead, “Fundamental limitations in
microelectronics—I, MOS technology,” Solid State Electron., vol. 15,
no. 7, pp. 819–829, Jul. 1972.
[8] R. W. Keyes, “Physical limits in digital electronics,” Proc. IEEE, vol. 63,
no. 5, pp. 740–767, May 1975.
[9] X. Tang, V. K. De, and D. Meindl, “Intrinsic MOSFET parameter fluctu-
ations due to random dopant placement,” IEEE Trans. Very Large Scale
Integr. (VLSI) Syst., vol. 5, no. 4, pp. 369–376, Dec. 1997.
[10] K. Lakshmikumar, R. Hadaway, and M. Copeland, “Characterization
and modeling of mismatch in MOS transistors for precision analog de-
sign,” IEEE J. Solid-State Circuits, vol. SSC-21, no. 6, pp. 1057–1066,
Dec. 1986.
Fig. 14. Comparison of the Vth distributions for low and high Vds for a [11] K. Takeuchi, T. Tatsumi, and A. Furukawa, “Channel engineering for the
sample set of 100 000 MOSFETs. reduction of random-dopant-placement-induced threshold voltage fluctu-
ation,” in IEDM Tech. Dig., 1997, pp. 841–844.
[12] S. Toriyama, D. Hagishima, K. Matsuzawa, and N. Sano, “Device simula-
tion of random dopant effects in ultra-small MOSFETs based on advanced
physical models,” in Proc. SISPAD, 2006, pp. 111–114.
[13] D. Frank, Y. Taur, M. Leong, and H. Wong, “Monte Carlo modeling of
threshold variation due to dopant fluctuations,” in VLSI Symp. Tech. Dig.,
1999, pp. 169–170.
[14] S. Roy, A. Lee, A. R. Brown, and A. Asenov, “Applicability of quasi-
3D and 3D MOSFET simulations in the ‘atomistic’ regime,” Kluwer J.
Comput. Electron., vol. 2, no. 2–4, pp. 423–426, Dec. 1998.
[15] S. Nobuyuki and T. Masaaki, “Random dopant model for three-
Fig. 15. Transistor cartoon useful for the proofs in the Appendix. dimensional drift-diffusion simulations in metal oxide semiconductor
field-effect-transistors,” Appl. Phys. Lett., vol. 79, no. 14, pp. 2267–2269,
Oct. 2001.
Calculating the integral and taking the limit of this func- [16] J. Yin, X. Shi, and R. Huang, “A new method to simulate random dopant
2
tion to infinity, we find limρ→∞ E[Φi ] = (qNA /2εSi )Wdm = induced threshold voltage fluctuations in sub-50 nm MOSFETSs with
non-uniform channel doping,” Solid State Electron., vol. 50, no. 9/10,
(kT /2) ln(NA /ni ) = 2ψB , where ψB is the Fermi potential. pp. 1551–1556, Sep./Oct. 2006.
In a similar manner for the electric field, we can write [17] R. W. Keyes, “Effect of randomness in the distribution of impurity ions
on FET thresholds in integrated electronics,” IEEE J. Solid-State Circuits,
NA 
N vol. SSC-10, no. 4, pp. 245–247, Aug. 1975.
E[Ez ] = bir [18] P. A. Stolk, F. P. Widdershoven, and D. B. M. Klaasen, “Modeling sta-
a3 r=1 tistical dopant fluctuations in MOS transistors,” IEEE Trans. Electron
⎡W Devices, vol. 45, no. 9, pp. 1960–1971, Sep. 1998.
ρ 2π dm [19] X. Tang, K. A. Bowman, J. C. Eble, V. K. De, and J. D. Meindl, “Impact
∼ qNA ⎣ rdrdϕdz(z + b) of random dopant placement on CMOS delay and power dissipation,” in
= Proc. 29th ESSDERC, 1999, pp. 184–187.
4πεSi [r2 + (z + b)2 ]3/2 [20] P. A. Stolk and D. B. M. Klaasen, “The effect of statistical dopant
0 0 0
⎤ fluctuations on MOS device performance,” in IEDM Tech. Dig., 1996,
Tox pp. 627–630.
−rdrdϕdz(z + b) ⎦ [21] K. Takeuchi, “Channel size dependence of dopant-induced threshold volt-
+ . age fluctuations,” in VLSI Symp. Tech. Dig., 1998, pp. 72–73.
[r2 + (z + b)2 ]3/2 [22] D. Reid, C. Millar, G. Roy, S. Roy, and A. Asenov, “Analysis of thresh-
−Wdm −Tox
old voltage distribution due to random dopants: A 100 000-sample 3-D
simulation study,” IEEE Trans. Electron Devices, vol. 56, no. 10,
After integration and taking the limit of this function to in- pp. 2255–2263, Oct. 2009.
finity, we have limρ→∞ E[EZ ] = qNA Wdm /εSi . Note that this [23] G. Panagopoulos and K. Roy, “A physical 3-D analytical model for the
expression represents the well-known formula (1-D analysis) threshold voltage considering RDF,” in Proc. DRC, 2009, pp. 81–82.
[24] S. Mukhopadhyay, K. Kim, K. A. Jenkins, C.-T. Chuang, and
for the surface and electric field. K. Roy, “An on-chip test structure and digital measurement method
for statistical characterization of local random variability in a
process,” IEEE J. Solid-State Circuits, vol. 43, no. 9, pp. 1951–1963,
R EFERENCES Sep. 2008.
[1] H. Wong and Y. Taur, “Three-dimensional atomistic simulation of discrete [25] Y. Li and C.-H. Hwang, “High-frequency characteristic fluctuations of
dopant distribution effects in sub-0.1 μm MOSFETs,” in IEDM Tech. nano-MOSFET circuit induced by random dopants,” IEEE Trans. Microw.
Dig., 1993, pp. 705–708. Theory Tech., vol. 56, no. 12, pp. 2726–2733, Dec. 2008.
[2] A. Asenov, “Random dopant induced threshold voltage lowering and fluc- [26] J. Yeung and H. Mahmoodi, “Robust sense amplifier design under random
tuations in sub-0.1 μm MOSFETs: A 3-D ‘atomistic’ simulation study,” dopant fluctuations in nanoscale CMOS technologies,” in Proc. SOC
IEEE Trans. Electron Devices, vol. 45, no. 12, pp. 2505–2513, Dec. 1998. Conf., Sep. 2006, pp. 261–264.
[3] Y. Li, S. Yu, J. Hwang, and F. Yang, “Discrete dopant fluctuations in [27] D. Burnett, K. Erington, C. Subramanian, and K. Baker, “Implications
20-nm/15 nm-gate planar CMOS,” IEEE Trans. Electron Devices, vol. 55, of fundamental threshold voltage variations for high-density SRAM and
no. 6, pp. 1449–1455, Jun. 2008. logic circuits,” in VLSI Symp. Tech. Dig., Jun. 1994, pp. 15–16.
PANAGOPOULOS AND ROY: ANALYTICAL MODEL FOR RDF-INDUCED THRESHOLD VOLTAGE VARIATIONS
[28] Y. Li, C.-H. Hwang, T.-C. Yeh, and T.-Y. Li, “Large-scale atomistic ap-
proach to random-dopant-induced characteristic variability in nanoscale
CMOS digital and high-frequency integrated circuits,” in Proc. ICCAD,
Nov. 2008, pp. 278–285.
[29] C. Millar, D. Reid, G. Roy, S. Roy, and A. Asenov, “Accurate statisti-
cal description of random dopant-induced threshold voltage variability,”
IEEE Trans. Electron Device Lett., vol. 29, no. 8, pp. 946–948, Aug. 2008.
[30] X.-W. Jiang, H.-X. Deng, J.-W. Luo, S.-S. Li, and L.-W. Wang, “A
fully three-dimensional atomistic quantum mechanical study on random
dopant-induced effects in 25-nm MOSFETs,” IEEE Trans. Electron De-
vices, vol. 55, no. 7, pp. 1720–1726, Jul. 2008.
[31] A. Asenov, G. Slavcheva, A. R. Brown, J. H. Davies, and S. Saini,
“Increase in the random dopant induced threshold fluctuations and low-
ering in sub-100 nm MOSFETs due to quantum effects: A 3-D density-
gradient simulation study,” IEEE Trans. Electron Devices, vol. 48, no. 4,
pp. 722–729, Apr. 2001.
[32] S. Sze and K. Ng, Physics of Semiconductor Devices, 3rd ed. Hoboken,
NJ: Wiley, 2008.
[33] R. F. Pierret, Fundamental of Solid State Devices, 2nd ed: Addition-
Wesley, 1996.
[34] Y. Taur and T. Ning, Fundamentals of Modern VLSI Devices.
Cambridge, U.K.: Cambridge Univ. Press, 1998.
[35] A. Papoulis and S. U. Pillai, Probability, Random Variables and Stochastic
Processes, 4th ed. New York: McGraw-Hill, 2002.
[36] C. A. Balanis, Advanced Engineering Electromagnetics. Hoboken, NJ:
Wiley, 1989.
[37] S. Ramo, J. R. Whinnery, and T. van Duzer, Fields and Waves in Commu-
nication Electronics, 2nd ed. Hoboken, NJ: Wiley, 1984.
Georgios Panagopoulos (S’05) received the
Diploma (with honors) degree in computer and
communication engineering from the University of
Thessaly, Volos, Greece, in 2006. He is working
toward the Ph.D. degree in electrical and computer
engineering with the School of Electrical and
Computer Engineering, Purdue University, West
Lafayette, IN.
From 2006 to 2007, he was a Research Assis-
tant designing analog circuits with the University
of Thessaly, Greece. His current research interests
include spintronics, device/circuit co-design for nanoscaled low-power and
reliable circuits, and design of process-variation-tolerant VLSI circuits and
systems.
Mr. Panagopoulos is the recipient of Academic Excellence Awards and
several scholarships during his studies.
403
Kaushik Roy (S’83–M’83–SM’95–F’02) received
the B.Tech. degree in electronics and electrical com-
munications engineering from the Indian Institute of
Technology, Kharagpur, India, and the Ph.D. degree
from the electrical and computer engineering from
the University of Illinois, Urbana, in 1990.
He was with the Semiconductor Process and De-
sign Center, Texas Instruments Incorporated, Dallas,
where he worked on field-programmable gate array
architecture development and low-power circuit de-
sign. He joined the electrical and computer engineer-
ing faculty at Purdue University, West Lafayette, IN, in 1993, where he is
currently a Professor and holds the Roscoe H. George Chair of Electrical &
Computer Engineering. He supervised 50 Ph.D. students. He has published
more than 500 papers in refereed journals and conference proceedings. He is
a coauthor of two books on low power CMOS VLSI design (John Wiley &
McGraw Hill). He is the holder of 15 patents. His research interests include
spintronics, VLSI design/computer-aided design for nanoscale silicon and
nonsilicon technologies, low-power electronics for portable computing and
wireless communications, VLSI testing and verification, and reconfigurable
computing.
Dr. Roy was a Research Visionary Board Member of Motorola Labo-
ratories in 2002 and held the M.K. Gandhi Distinguished Visiting faculty
at Indian Institute of Technology (Bombay). He has been in the Editorial
Board of IEEE Design and Test, IEEE TRANSACTIONS ON CIRCUITS and
SYSTEMS, and IEEE TRANSACTIONS ON VLSI SYSTEMS. He was a Guest
Editor for Special Issue on Low-Power VLSI in the IEEE Design and Test
(1994) and IEEE TRANSACTIONS ON VLSI SYSTEMS (June 2000), IEEE
Proceedings–Computers and Digital Techniques (July 2002). He was the
recipient of the National Science Foundation Career Development Award in
1995; IBM Faculty Partnership Award; ATT/Lucent Foundation Award; 2005
SRC Technical Excellence Award; SRC Inventors Award; Purdue College of
Engineering Research Excellence Award; Humboldt Research Award in 2010;
best paper awards at the 1997 International Test Conference, IEEE 2000
International Symposium on Quality of IC Design, 2003 IEEE Latin American
Test Workshop, 2003 IEEE Nano, 2004 IEEE International Conference on
Computer Design, 2006 IEEE/ACM International Symposium on Low Power
Electronics & Design, and 2005 IEEE Circuits; System Society Outstanding
Young Author Award (Chris Kim); and the 2006 IEEE TRANSACTIONS ON
VLSI SYSTEMS Best Paper Award. He is a Purdue University Faculty Scholar.