scispace - formally typeset
Search or ask a question
Journal ArticleDOI

Adiabatic Superconducting Artificial Neural Network: Basic Cells.

TL;DR: In this article, the authors considered adiabatic superconducting cells operating as an artificial neuron and synapse of a multilayer perceptron and their compact circuits contain just one and two Josephson junctions, respectively.
Abstract: We consider adiabatic superconducting cells operating as an artificial neuron and synapse of a multilayer perceptron (MLP). Their compact circuits contain just one and two Josephson junctions, respectively. While the signal is represented as magnetic flux, the proposed cells are inherently nonlinear and close-to-linear magnetic flux transformers. The neuron is capable of providing a one-shot calculation of sigmoid and hyperbolic tangent activation functions most commonly used in MLP. The synapse features by both positive and negative signal transfer coefficients in the range ~ (-0.5,0.5). We briefly discuss implementation issues and further steps toward multilayer adiabatic superconducting artificial neural network which promises to be a compact and the most energy-efficient implementation of MLP.

Summary (1 min read)

Introduction

  • The authors consider adiabatic superconducting cells operating as an artificial neuron and synapse of a multilayer perceptron (MLP).
  • The synapse features both positive and negative signal transfer coefficients in the range ( 0:5, 0:5).
  • Artificial neural network (ANN) is the key technology in the fast developing area of artificial intelligence.
  • It has been already broadly introduced in their everyday life.
  • Besides the several attempts to the implementation of the superconducting ANNs proposed since the 1990s,4–12 the idea to adopt the adiabatic logic cells to neuromorphic circuits was presented only recently.

II. BASIC CELLS

  • The basic element of superconducting circuits is the Josephson junction.
  • Contrary to semiconductor transistor, the Josephson junction is not fabricated in a substrate but between two superconductor layers deposited on a substrate utilized as a mechanical support.
  • In most cases, a configurable ANN would be preferable.
  • This phenomenon was already proposed for utilization in artificial synapse of superconducting spiking ANN.
  • With the fixed critical currents, the shape of the transfer function is determined by inductances lin, l.

III. DISCUSSION

  • Both considered cells operate in a pure superconducting regime.
  • Nevertheless, the proposed adiabatic superconducting ANN can be up to 104–105 times more energy efficient than its semiconductor counterparts.
  • 1,26 However, such aspects as the linearity of amplification, the distance of signal propagation without amplification, and related issues of achievable fan-in and fan-out should be additionally considered.
  • Another feature is the periodicity of sigma-cell based neuron transfer function.
  • In particular case of the proposed synapse, one could benefit from implementation of the magnetic Josephson junction controlled by direction of magnetic field, like the Josephson magnetic rotary valve30 with heterogeneous area of weak link.

IV. CONCLUSION

  • The authors considered operation principles of adiabatic superconducting basic cells for implementation of multilayer perceptron.
  • These are artificial neuron and synapse which are nonlinear and close-to-linear superconducting transformers of magnetic flux, respectively.
  • Both cells are capable of operation in the adiabatic regime featured by ultra-low power consumption at the level of 4 to 5 orders of magnitude less than that of their modern semiconductor counterparts (including cooling power penalty).
  • The synapse is implemented with two magnetic Josephson junctions with controllable critical currents.
  • It provides both positive and negative signal transfer coefficients in the range ( 0:5, 0:5).

Did you find this useful? Give us your feedback

Content maybe subject to copyright    Report

Adiabatic superconducting artificial neural network: Basic cells
Igor I. Soloviev, Andrey E. Schegolev, Nikolay V. Klenov, Sergey V. Bakurskiy, Mikhail Yu. Kupriyanov, Maxim V.
Tereshonok, Anton V. Shadrin, Vasily S. Stolyarov, and Alexander A. Golubov
Citation: Journal of Applied Physics 124, 152113 (2018); doi: 10.1063/1.5042147
View online: https://doi.org/10.1063/1.5042147
View Table of Contents: http://aip.scitation.org/toc/jap/124/15
Published by the American Institute of Physics
Articles you may be interested in
Theoretical basis of SQUID-based artificial neurons
Journal of Applied Physics 124, 152106 (2018); 10.1063/1.5037718
Chain of magnetic tunnel junctions as a spintronic memristor
Journal of Applied Physics 124, 152116 (2018); 10.1063/1.5042431
Ultra-fast logic devices using artificial “neurons” based on antiferromagnetic pulse generators
Journal of Applied Physics 124, 152115 (2018); 10.1063/1.5042348
Overcoming device unreliability with continuous learning in a population coding based computing system
Journal of Applied Physics 124, 152111 (2018); 10.1063/1.5042250
Integrate-and-fire neuron circuit using positive feedback field effect transistor for low power operation
Journal of Applied Physics 124, 152107 (2018); 10.1063/1.5031929
Tutorial: Fabrication and three-dimensional integration of nanoscale memristive devices and arrays
Journal of Applied Physics 124, 152001 (2018); 10.1063/1.5038109

Adiabatic superconducting articial neural network: Basic cells
Igor I. Soloviev,
1,2,3,a)
Andrey E. Schegolev,
1,2,4,5
Nikolay V. Klenov,
1,2,4,5,6
Sergey V. Bakurskiy,
1,2,3
Mikhail Yu. Kupriyanov,
1
Maxim V. Tereshonok,
2,5
Anton V. Shadrin,
3
Vasily S. Stolyarov,
3,6,7,8
and Alexander A. Golubov
3,9
1
Lomonosov Moscow State University Skobeltsyn Institute of Nuclear Physics, 119991 Moscow, Russia
2
MIREARussian Technological University, 119454 Moscow, Russia
3
Moscow Institute of Physics and Technology, 141700 Dolgoprudny, Russia
4
Physics Department, Lomonosov Moscow State University, 119991 Moscow, Russia
5
Moscow Technical University of Communications and Informatics (MTUCI), 111024 Moscow, Russia
6
Dukhov Research Institute of Automatics (VNIIA), 127055 Moscow, Russia
7
Institute of Solid State Physics RAS, 142432 Chernogolovka, Russia
8
Solid State Physics Department, KFU, 420008 Kazan, Russia
9
Faculty of Science and Technology and MESA+ Institute of Nanotechnology, 7500 AE Enschede,
The Netherlands
(Received 30 May 2018; accepted 25 July 2018; published online 26 September 2018)
We consider adiabatic superconducting cells operating as an articial neuron and synapse of a multi-
layer perceptron (MLP). Their compact circuits contain just one and two Josephson junctions,
respectively. While the signal is represented as magnetic ux, the proposed cells are inherently non-
linear and close-to-linear magnetic ux transformers. The neuron is capable of providing the one-
shot calculation of sigmoid and hyperbolic tangent activation functions most commonly used in
MLP. The synapse features both positive and negative signal transfer coefcients in the range
(0:5, 0:5). We briey discuss implementation issues and further steps toward the multilayer adi-
abatic superconducting articial neural network, which promises to be a compact and the most
energy-efcient implementation of MLP. Published by AIP Publishing.
https://doi.org/10.1063/1.5042147
I. INTRODUCTION
Articial neural network (ANN) is the key technology in
the fast developing area of articial intelligence. It has been
already broadly introduced in our everyday life. Further pro-
gress requires an increase in complexity and depth of ANNs.
However, modern implementations of the neural networks
are commonly based on conventional computer hardware
which is not well suited for neuromorphic operation. This
leads to excessive power consumption and hardware over-
head. Ideal basic elements of ANNs should combine the mul-
tiple properties like one-shot calculation of their functions,
operation with energy near the thermal noise oor, and nano-
scale dimensions.
The most energy ef cient computing today can be per-
formed using the superconductor digital technology.
1
The rst ever practical logic gates capable of operating down
to and below the Landauer thermal limit
2
were realized
recently
3
on the basis of adiabatic superconductor logic.
Besides the several attempts to the implementation of the
superconducting ANNs proposed since the 1990s,
412
the
idea to adopt the adiabatic logic cells to neuromorphic cir-
cuits was presented only recently.
13,14
In this paper, we con-
sider operation principles of adiabatic superconducting basic
cells which comply with the above-mentioned properties for
ANN implementation. We focus on a particular multilayer
perceptron (MLP) because of a wide range of its applicability
and well-developed learning algorithms for such a network.
II. BASIC CELLS
The basic element of superconducting circuits is the
Josephson junction. Its characteristic energy typically lies
below aJ level while switching frequency is several hundred
GHz. Contrary to semiconductor transistor, the Josephson
junction is not fabricated in a substrate but between two
superconductor layers deposited on a substrate utilized as a
mechanical support. This provides opportunity for supercon-
ducting circuits to benet from 3D topology which can be
especially suitable for deep ANNs. The minimal feature size
of superconducting circuits is progressively decreased down
to nanoscales in recent years.
15
Another attractive feature of the Josephson junction is its
inherently strong nonlinearity. Indeed, the current owing
through the junction, I, is commonly related to the supercon-
ducting phase difference between the superconducting banks,
w
,as
I ¼ I
c
sin
w
, (1)
where I
c
is the junction critical current. We show below that
this current-phase relation (CPR) having both linear and non-
linear parts is well suited for implementation of supercon-
ducting arti cial neuron with one-shot calculation of sigmoid
or hyperbolic tangent activation functions
σ (x) ¼
1
1 þ e
x
, (2a)
a)
isol@phys.msu.ru
JOURNAL OF APPLIED PHYSICS 124, 152113 (2018)
0021-8979/2018/124(15)/152113/5/$30.00 124, 152113-1 Published by AIP Publishing.

or
τ (x) ¼ tanh (x), (2b)
utilized in MLP and superconducting synapse enabling
signal transfer with both positive and negative coefcients.
Unlike most of their predecessors,
49,11,12
both cells are oper-
ating in a pure superconducting mode featured by minimal
power consumption.
A. Articial neuron
One of the simplest superconducting cells is parametric
quantron proposed in 1982 for adiabatic operation.
16
It is the
superconducting loop consisted of a Josephson junction and
a superconducting inductance. According to the Josephson
junction CPR (1), the relation between the input magnetic
ux and the Josephson junction phase in its circuit has a
simple expression:
w
þ l sin
w
¼
f
in
, (3)
where we use normalization of current to critical current of
the Josephson junction, I
c
, and input magnetic ux Φ
in
to the
magnetic ux quantum Φ
0
,
f
in
¼ 2πΦ
in
=Φ
0
, inductance, L,
is normalized to characteristic inductance, l ¼ L=L
c
,
L
c
¼ Φ
0
=2πI
c
, accordingly.
It is seen from (1) and (3) that the current circulating in
the loop has a tilted sine dependence on input magnetic ux.
The way to transform this dependence close to the desired
one [(2a) or (2b)] is the addition of a linear term compensat-
ing the sine slope on the initial section (where sin
w
w
)in
the vicinity of zero input ux,
f
in
0.
This can be done by attaching another superconducting
loop with a part of its inductance, l
out
, being common with
the initial circuit [see Fig. 1(a)]. The synthesized cell was
named a sigma-cell
13
because its transformation of mag-
netic ux can be very close to sigmoid function. Here, we
are interested in a transfer function,
f
out
(
f
in
), where output
magnetic ux,
f
out
, is proportional to output current,
f
out
¼ l
out
i
out
.
The system of equations describing the proposed cell is
as follows:
w
þ l sin
w
¼
f
in
=2 þ l
out
i
out
, (4a)
w
þ l sin
w
¼
f
in
þ l
a
i
a
, (4b)
where l
a
is the attached inductance. The corresponding
system implicitly dening the transfer function through
dependencies of
f
out
,
f
in
on
w
has the following form:
f
out
¼ l
out
f
in
2l
a
sin
w
2(l
a
þ l
out
)
, (5a)
f
in
¼ 2
l
a
þ l
out
l
a
þ 2l
out

w
þ l þ
l
a
l
out
l
a
þ l
out

sin
w

: (5b)
Vanishing of the derivative d
f
out
=d
f
in
at
f
in
¼ 0 corre-
sponds to the condition:
l
a
¼ 1 þ l: (6)
One can t (5) to sigmoid function (2a) taking (6) into
account with the two tting parameters: l, l
out
.
The result of tting is shown in Fig. 1(b). The found
optimal values, l ¼ 0:125, l
out
¼ 0:3, provide conformity of
the sigma-cell transfer function with sigmoid one with stan-
dard deviation at the level of 10
3
. Sigmoid function (2a)
was scaled as σ (1:173x) in our tting process. The transfer
function
f
out
(
f
in
) (5) was normalized by 2πl
out
=(l
a
þ 2l
out
)
to t a unit height and shifted by a half period. The latter can
be obtained by application of a constant bias ux to the
circuit,
f
b
¼2π(l
a
þ l
out
)=(l
a
þ 2l
out
).
While sigmoid activation function is commonly used for
input data dened in the positive domain, for data dened on
the whole numeric axis around zero, it is convenient to use
hyperbolic tangent. Application of additional bias ux provid-
ing π phase shift into the loop containing Josephson junction
moves the center of the nonlinear part of the cell transfer func-
tion to zero. This allows one to obtain the desired shape of
activation function (2b).Theπ phase shift can also be imple-
mented using the πJosephson junction
1720
with π shift of its
CPR (1), I ¼I
c
sin (
w
), instead of the standard one.
One needs to correspondingly change the sign of the
terms containing sine function in (5) to perform the tting
FIG. 1. (a) Scheme of an articial neuron cell. (b) The cell transfer function
(line) tted to sigmoid and hyperbolic tangent functions (dots). Scaling of
the functions (2) is shown in the gure. The transfer function
f
out
(
f
in
)is
normalized by 2πl
out
=(l
a
þ 2l
out
) and shifted by 2π(l
a
þ l
out
)=(l
a
þ 2l
out
)
on the ux axis to t (2a), and normalized to πl
out
=(l
a
þ 2l
out
) with no addi-
tional shift on ux axis to t (2b). The optimal values of parameters are
l ¼ 0:125, l
out
¼ 0:3, l
a
¼ 1:125. Consistency of curves in both cases is at
the level of 10
3
. Hyperbolic tangent activation function is tted with π shift
in the Josephson junction CPR (1).
152113-2 Soloviev et al. J. Appl. Phys. 124, 152113 (2018)

procedure. The tting result is presented in Fig. 1(b).
Hyperbolic tangent function was scaled as tanh (0:586x)
while the transfer function
f
out
(
f
in
) was normalized by a
factor of two lower value than the previous time,
πl
out
=(l
a
þ 2l
out
). With the same values of parameters l, l
out
,
and zero bias ux, we obtained the same conformity of the
curves.
B. Articial synapse
Synapse modulates the weight of a signal arriving at
the neuron. In our case, the signal corresponds to magnetic
ux and, therefore, synapse can be implemented simply as a
transformer of magnetic ux with desired coupling factor.
Summation of signals can be provided by connecting the
transformers to a single superconducting input loop of the
neuron. However, this solution suits for ANN with a certain
and unchangeable conguration.
In most cases, a congurable ANN would be preferable.
The selected conguration of inter-neuron connections
should be maintained during its entire use if the feature space
dimensions do not vary. However, the weight values should
be congurable if we want to train the ANN on the y. The
best way to meet this requirement is utilization of some non-
volatile memory elements. In superconducting circuits, such
an element can be implement ed by using the ferromagnetic
(F) materials. In particular, introduction of F-layers into the
Josephson junction weak link area allows us to modulate its
critical current.
1,21,22
This phenomenon was already proposed
for utilization in articial synapse of superconducting spiking
ANN.
12
In our case of MLP, we can also make use of it.
The synapse scheme presented in Fig. 2(a) is nearly a
mirrored scheme of the proposed neuron [Fig. 1(a)]. The
only differences are the addition of the second Josephson
junction and the possibility to independently modulate criti-
cal currents of the magnetic junctions (marked by boxes),
e.g., by application of tuning magnetic eld.
For MLP, it is required to provide both positive and
negative weights of signal. Our synapse is designed accord-
ing to this requirement. The input current, i
in
, induced in
inductance l
in
by input magnetic ux,
f
in
, is split toward
the two Josephson junctions. Magnitude of currents i
1
, i
2
in each branch correspond to critical currents of the junc-
tions, i
c1
, i
c2
, so that the sign of output circulating current,
i
cir
¼ (i
1
i
2
)=2 (and the direction of output magnetic ux,
f
out
), is determined by their ratio. Maximum in equality of
i
c1
, i
c2
provides maximum output signal, while equal critical
currents correspond to zero transfer coefcient.
It is convenient to present the system of equations for
the synapse cell in terms of Josephson junctions phase sum,
w
þ
¼ (
w
1
þ
w
2
)=2, and phase difference,
w
¼ (
w
1
w
2
)=2:
w
þ
þ
l
2
þ l
in

i
in
þ
f
in
¼ 0, (7a)
w
þ li
cir
¼ 0: (7b)
Furthermore, introducing the sum Σi
c
¼ i
c1
þ i
c2
and differ-
ence Δi
c
¼ i
c1
i
c2
of the critical currents and taking (1) into
account one can represent (7) in the following form:
w
þ
þ
l
2
þ l
in

ðΣi
c
sin
w
þ
cos
w
þ Δi
c
sin
w
cos
w
þ
Þ
þ
f
in
¼ 0;
(8a)
w
þ
l
2
(Σi
c
sin
w
cos
w
þ
þ Δi
c
sin
w
þ
cos
w
) ¼ 0: (8b)
The dependence of the phase difference on the phase sum,
w
(
w
þ
), can be obtained
23,24
from (8b) with corresponding
function
f (
w
,
w
þ
) ¼
w
þ
l
2
(Σi
c
sin
w
cos
w
þ
þ Δi
c
sin
w
þ
cos
w
),
(9)
as follows:
w
¼
ð
πsgnΔi
c
0
H[ f (x;
w
þ
)sgn Δ i
c
]dx; (10)
where H(x) is the Heaviside step function. Equations (7a),
(8a), and (10) implicitly dene the cell transfer function
FIG. 2. (a) Scheme of an articial synapse cell. Magnetic Josephson junc-
tions are marked by boxes. (b) Synapse cell transfer function for the values
of parameters: l
in
¼ 2, l ¼ 4, Σi
c
¼ 1, and Δi
c
as shown in the gure.
Vertical dotted line shows the boundary of highly linear range where stan-
dard deviation from the linear function is at the level of 10
3
. This range
corresponds to maximum output magnetic ux of the optimized neuron cell.
152113-3 Soloviev et al. J. Appl. Phys. 124, 152113 (2018)

f
out
(
f
in
) through dependencies
f
out
¼ 2li
cir
¼2
w
(
w
þ
)
and
f
in
[
w
(
w
þ
),
w
þ
]on
w
þ
. Here, we are interested in the
range of the phase sum,
w
þ
[ [0, π=2), where the transfer
function might be linear.
Figure 2(b) shows synapse cell transfer function for dif-
ferent values of critical currents difference in the range
Δi
c
[ [ 0:9, 0:9]. The critical current sum is Σi
c
¼ 1. With
the xed critical currents, the shape of the transfer function is
determined by inductances l
in
, l.
In accordance with (7a), an increase in input inductance
l
in
increases the amplitude of nonlinearity of the dependence
of input current on input ux i
in
(
f
in
) making it more tilted.
This is in complete analogy with parametric quantron
scheme (3). The slope of the linear part of the transfer func-
tion is correspondingly decreased. However, this gives a
stretching of this linear part, which is of use for us, and con-
traction of the nonlinear part.
Increase in inductance l provides the same effect [see
(7a)]. At the same time, it increases the nonlinearity of the
dependence of output ux on phase sum [see (8b)] which
vice versa increases the slope of the linear part though
making it less linear. The goal of optimization of the transfer
function
f
out
(
f
in
) is the maximum modulation of its slope
alongside with the high linearity among the possibly wider
range of input ux.
In our case, the values of inductances were chosen to be
l
in
¼ 2, l ¼ 4. With these parameters magnetic ux can be
transferred through the synapse with coefcients in the range
(0:5, 0:5) depending on the critical currents difference.
For maximum output magnetic ux of optimized neuron,
2πl
out
=(l
a
þ l
out
) 1:1, maximum standard deviation of the
synapse transfer function from the linear function is at the
level of 10
3
. In the whole shown range [0, π], it is of an
order of magnitude worse.
III. DISCUSSION
Both considered cells operate in a pure superconducting
regime. Evolution of their states is fully physically reversible.
Therefore, they can be operated adiabatically with energy per
operation down to the Landauer limit.
2
For standard working
temperature of superconducting circuits, T ¼ 4:2 K, this limit
corresponds to the energy, k
B
T ln 2 4 10
23
J (where k
B
is the Boltzmann constant). Estimations show that the bit
energy can be as low as 10
21
J for adiabatic superconductor
logic at clock frequency of 10 GHz.
25
This is million times
less than characteristic energy consumed by a semiconductor
transistor. In one hand, taking into account the fact that
modern implementation of neuron based on complementary-
metal-oxide semiconductor (CMOS) technology requires a
few dozens of transistors, the possible gap between power
consumption of semiconductor and superconductor ANN is
increased by an order. On the other hand, penalty for super-
conducting circuits cooling is typically several hundred W/W
that cancels out the two to three orders of supremacy.
Nevertheless, the proposed adiabatic superconducting ANN
can be up to 10
4
10
5
times more energy efcient than its
semiconductor counterparts.
One should note some peculiarities of the proposed
concept. First of all, there is no power supply in these circuits
and so the signal vanishes. Therefore, there is a need for a
ux amplier which can be implemented on a base of some
standard adiabatic cell like adiabatic quantum ux parame-
tron (AQFP).
1,26
However, such aspects as the linearity of
amplication, the distance of signal propagation without
amplication, and related issues of achievable fan-in and
fan-out should be additionally considered.
Another feature is the periodicity of sigma-cell based
neuron transfer function. Corresponding issues can be miti-
gated by a signal normalization.
Along with the use of standard superconducting inte-
grated circuits fabrication process, the proposed cells require
utilization of magnetic Josephson junctions which are rela-
tively new to superconducting technology. Nevertheless,
modern developments of cryogenic magnetic memory
1,27
and superconducting logic circuits with controlled functional-
ity
28,29
promise their fast introduction.
In particular case of the proposed synapse, one could
benet from implementation of the magnetic Josephson junc-
tion controlled by direction of magnetic eld, like the
Josephson magnetic rotary valve
30
with heterogeneous area
of weak link. Such a valve is featured by high critical current
for a certain direction of its F-layer magnetization and low
critical current for the direction rotated by 90
. Two such
junctions in close proximity to each other with mutual rota-
tion on 90
relative to their axes directed along the boundary
of inhomogeneity allow one to obtain high critical current for
one junction and low critical current for another one with the
same direction of magnetizations of their F-layers. In this
case, rotation of their magnetizations leads to a correspond-
ing decrease and increase of Josephson junctions critical
currents which means modulation of synapse weight, accord-
ing to Fig. 2. Utilization of the rotary valve reduces the
number of control lines required to program the magnetic
Josephson junctions by half. However, their total number,
which is twice the number of synapses, remains huge for
practical ANNs. Therefore, the effective synapse control
is another urgent task on the way to multilayer adiabatic
superconducting ANN.
IV. CONCLUSION
In this paper, we considered operation principles of
adiabatic superconducting basic cells for implementation of
multilayer perceptron. These are articial neuron and synapse
which are nonlinear and close-to-linear superconducting trans-
formers of magnetic ux, respectively. Both cells are capable
of operation in the adiabatic regime featured by ultra-low
power consumption at the level of 4 to 5 orders of magnitude
less than that of their modern semiconductor counterparts
(including cooling power penalty). The proposed neuron cell
contains just a single Josephson junction. The neuron provides
one-shot calculation of either sigmoid or hyperbolic tangent
activation function. The certain type of this function is deter-
mined by the type of utilized Josephson junction and can also
be switched on the y by application of magnetic ux. The
synapse is implemented with two magnetic Josephson
152113-4 Soloviev et al. J. Appl. Phys. 124, 152113 (2018)

Citations
More filters
Journal ArticleDOI
TL;DR: This work reviews the work that has been done on biologically inspired superconductive devices, circuits, and architectures and discusses the scaling potential of these demonstrations.
Abstract: Neuromorphic computing is a broad field that uses biological inspiration to address computing design. It is being pursued in many hardware technologies, both novel and conventional. We discuss the use of superconductive electronics for neuromorphic computing and why they are a compelling technology for the design of neuromorphic computing systems. One example is the natural spiking behavior of Josephson junctions and the ability to transmit short voltage spikes without the resistive capacitive time constants that typically hinder spike-based computing. We review the work that has been done on biologically inspired superconductive devices, circuits, and architectures and discuss the scaling potential of these demonstrations.

27 citations

Journal ArticleDOI
TL;DR: In this article, a periodic structure composed of ferromagnetic (F) layers spaced by thin superconductors (s) was proposed to check the feasibility of controllable switching between AP and P states through the whole periodic structure.
Abstract: We present a study of magnetic structures with controllable effective exchange energy for Josephson switches and memory applications. As a basis for a weak link we propose to use a periodic structure composed of ferromagnetic (F) layers spaced by thin superconductors (s). Our calculations based on the Usadel equations show that switching from parallel (P) to antiparallel (AP) alignment of neighboring F layers can lead to a significant enhancement of the critical current through the junction. To control the magnetic alignment we propose to use a periodic system whose unit cell is a pseudo spin valve of structure F1/s/F2/s where F1 and F2 are two magnetic layers having different coercive fields. In order to check the feasibility of controllable switching between AP and P states through the whole periodic structure, we prepared a superlattice [Co(1.5 nm)/Nb(8 nm)/Co(2.5 nm)/Nb(8 nm)]6 between two superconducting layers of Nb(25 nm). Neutron scattering and magnetometry data showed that parallel and antiparallel alignment can be controlled with a magnetic field of only several tens of Oersted.

26 citations

Journal ArticleDOI
TL;DR: In this article, a spin-polarized neutron reflectometry was used to investigate the magnetization profile of superlattices composed of ferromagnetic Gd and superconducting Nb layers.
Abstract: We have used spin-polarized neutron reflectometry to investigate the magnetization profile of superlattices composed of ferromagnetic Gd and superconducting Nb layers. We have observed a partial suppression of ferromagnetic (F) order of Gd layers in [$\mathrm{Gd}({d}_{F})/\mathrm{Nb}(25 \mathrm{nm}){\mathrm{]}}_{12}$ superlattices below the superconducting (S) transition of the Nb layers. The amplitude of the suppression decreases with increasing ${d}_{F}$. By analyzing the neutron spin asymmetry we conclude that the observed effect has an electromagnetic origin---the proximity-coupled S layers screen out the external magnetic field and thus suppress the F response of the Gd layers inside the structure. Our investigation demonstrates the considerable influence of electromagnetic effects on the magnetic properties of S/F systems.

18 citations

Journal ArticleDOI
TL;DR: Here I sketch a concept for optoelectronic hardware, beginning with synaptic circuits, continuing through wafer-scale integration, and extending to systems interconnected with fiber-optic white matter, potentially at the scale of the human brain and beyond.
Abstract: To design and construct hardware for general intelligence, we must consider principles of both neuroscience and very-large-scale integration. For large neural systems capable of general intelligence, the attributes of photonics for communication and electronics for computation are complementary and interdependent. Using light for communication enables high fan-out as well as low-latency signaling across large systems with no traffic-dependent bottlenecks. For computation, the inherent nonlinearities, high speed, and low power consumption of Josephson circuits are conducive to complex neural functions. Operation at 4\,K enables the use of single-photon detectors and silicon light sources, two features that lead to efficiency and economical scalability. Here I sketch a concept for optoelectronic hardware, beginning with synaptic circuits, continuing through wafer-scale integration, and extending to systems interconnected with fiber-optic white matter, potentially at the scale of the human brain and beyond.

17 citations

Journal ArticleDOI
TL;DR: In this paper, the authors explore the fan-in and fan-out in superconductive neuromorphic circuits based on Josephson junctions and demonstrate results in simulation at a level of 1-to-10,000 similar to the human brain.
Abstract: Neuromorphic computing has the potential to further the success of software-based artificial neural networks (ANNs) by designing hardware from a different perspective. Current research in neuromorphic hardware targets dramatic improvements to ANN performance by increasing energy efficiency, speed of operation, and even seeks to extend the utility of ANNs by natively adding functionality such as spiking operation. One promising neuromorphic hardware platform is based on superconductive electronics, which has the potential to incorporate all of these advantages at the device level in addition to offering the potential of near lossless communications both within the neuromorphic circuits as well as between disparate superconductive chips. Here we explore one of the fundamental brain-inspired architecture components, the fan-in and fan-out as realized in superconductive circuits based on Josephson junctions. From our calculations and WRSPICE simulations we find that the fan-out should be limited only by junction count and circuit size limitations, and we demonstrate results in simulation at a level of 1-to-10,000, similar to that of the human brain. We find that fan-in has more limitations, but a fan-in level on the order of a few 100-to-1 should be achievable based on current technology. We discuss our findings and the critical parameters that set the limits on fan-in and fan-out in the context of superconductive neuromorphic circuits.

13 citations

References
More filters
Journal ArticleDOI
TL;DR: In this paper, the authors studied the properties of S-F/N-sIS type Josephson junctions in the frame of the quasiclassical Usadel formalism.
Abstract: In this work, we study theoretically the properties of S-F/N-sIS type Josephson junctions in the frame of the quasiclassical Usadel formalism. The structure consists of two superconducting electrodes (S), a tunnel barrier (I), a combined normal metal/ferromagnet (N/F) interlayer, and a thin superconducting film (s). We demonstrate the breakdown of a spatial uniformity of the superconducting order in the s-film and its decomposition into domains with a phase shift π. The effect is sensitive to the thickness of the s layer and the widths of the F and N films in the direction along the sIS interface. We predict the existence of a regime where the structure has two energy minima and can be switched between them by an electric current injected laterally into the structure. The state of the system can be non-destructively read by an electric current flowing across the junction.

25 citations

Journal ArticleDOI
M Hidaka1, L A Akers1
TL;DR: A new type of artificial neural cell is described using superconducting circuits for implementing a weighted sum of inputs and the estimations show a promise for implementation of large scale neural networks.
Abstract: The authors describe a new type of artificial neural cell using superconducting circuits Circulating current flowing in a superconducting closed loop is used for implementing a weighted sum of inputs The weight value and sign are digitally variable The weight values are stored in superconducting memory loops, and can be controlled by an up-down counter The performance of the cell, such as cell size, power dissipation, and input/output capability, were evaluated and the estimations show a promise for implementation of large scale neural networks

24 citations

Journal ArticleDOI
TL;DR: In this paper, a scheme for the realization of artificial neural networks based on superconducting quantum interference devices (SQUIDs) was proposed, which can be particularly convenient as support for super-conducting applications such as detectors for astrophysics, high energy experiments, medicine imaging and so on.
Abstract: We propose a scheme for the realization of artificial neural networks based on superconducting quantum interference devices (SQUIDs). In order to demonstrate the operation of this scheme we designed and successfully tested a small network that implements an XOR gate and is trained by means of examples. The proposed scheme can be particularly convenient as support for superconducting applications such as detectors for astrophysics, high energy experiments, medicine imaging and so on.

23 citations

Journal ArticleDOI
TL;DR: In this article, a superconducting neural circuit is fabricated for the first time by use of a niobium integrated-circuit technology, where fluxon pulses on Josephson transmission lines (JTLs) are used as neural impulses.
Abstract: A superconducting neural circuit is fabricated for the first time by use of a niobium integrated‐circuit technology. Fluxon pulses on Josephson transmission lines (JTLs) are used as neural impulses. In this circuit a threshold element (a neuron) is composed of two JTL elements connected through a resistor. The conductance value of the resistor represent a synaptic strength. The fan‐in and the fan‐out are accomplished by the biased JTL branches. The operation of 2‐bit neural based A/D converter is successfully observed. These circuits do not require any hysteretic Josephson junctions, and hence, have a potential to be fabricated with the high‐Tc superconductors.

22 citations

Journal ArticleDOI
TL;DR: In this article, the effect of an external RF field on the switching processes of magnetic Josephson junctions (MJJs) suitable for the realization of fast, scalable cryogenic memories compatible with Single Flux Quantum logic was investigated.
Abstract: We test the effect of an external RF field on the switching processes of magnetic Josephson junctions (MJJs) suitable for the realization of fast, scalable cryogenic memories compatible with Single Flux Quantum logic We show that the combined application of microwaves and magnetic field pulses can improve the performances of the device, increasing the separation between the critical current levels corresponding to logical '0' and '1' The enhancement of the current level separation can be as high as 80% using an optimal set of parameters We demonstrate that external RF fields can be used as an additional tool to manipulate the memory states, and we expect that this approach may lead to the development of new methods of selecting MJJs and manipulating their states in memory arrays for various applications

21 citations

Frequently Asked Questions (17)
Q1. What are the two types of synapses?

These are artificial neuron and synapse which are nonlinear and close-to-linear superconducting transformers of magnetic flux, respectively. 

In this paper, the authors considered operation principles of adiabatic superconducting basic cells for implementation of multilayer perceptron. 

While sigmoid activation function is commonly used for input data defined in the positive domain, for data defined on the whole numeric axis around zero, it is convenient to use hyperbolic tangent. 

1,26 However, such aspects as the linearity of amplification, the distance of signal propagation without amplification, and related issues of achievable fan-in and fan-out should be additionally considered. 

The synthesized cell was named a “sigma-cell”13 because its transformation of magnetic flux can be very close to sigmoid function. 

the proposed adiabatic superconducting ANN can be up to 104–105 times more energy efficient than its semiconductor counterparts. 

At the same time, it increases the nonlinearity of the dependence of output flux on phase sum [see (8b)] which vice versa increases the slope of the linear part though making it less linear. 

In this paper, the authors considered operation principles of adiabatic superconducting basic cells for implementation of multilayer perceptron. 

2. Utilization of the rotary valve reduces the number of control lines required to program the magnetic Josephson junctions by half. 

Hyperbolic tangent function was scaled as tanh (0:586x) while the transfer function fout(fin) was normalized by a factor of two lower value than the previous time, πlout=(la þ 2lout). 

Two such junctions in close proximity to each other with mutual rotation on 90 relative to their axes directed along the boundary of inhomogeneity allow one to obtain high critical current for one junction and low critical current for another one with the same direction of magnetizations of their F-layers. 

In accordance with (7a), an increase in input inductance lin increases the amplitude of nonlinearity of the dependence of input current on input flux iin(fin) making it more tilted. 

In this case, rotation of their magnetizations leads to a corresponding decrease and increase of Josephson junction’s critical currents which means modulation of synapse weight, according to Fig. 

the authors are interested in a transfer function, fout(fin), where output magnetic flux, fout, is proportional to output current, fout ¼ loutiout. 

On the other hand, penalty for superconducting circuits cooling is typically several hundred W/W that cancels out the two to three orders of supremacy. 

The way to transform this dependence close to the desired one [(2a) or (2b)] is the addition of a linear term compensating the sine slope on the initial section (where sinw w) in the vicinity of zero input flux, fin 0. 

For standard working temperature of superconducting circuits, T ¼ 4:2 K, this limit corresponds to the energy, kBT ln 2 4 10 23 J (where kB is the Boltzmann constant).