I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t
201
9
,
p
p
.
2
4
6
9
~
2
4
8
0
I
SS
N:
2088
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
9
i
4
.
pp
2
4
6
9
-
2480
2469
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e
.
co
m/
jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
JE
C
E
Area a
nd
po
w
er e
ff
icien
t
VL
SI
a
rc
hitectu
re of
m
o
de
decisio
n
in
integer
m
o
tion
est
i
m
a
tion
for
H
EV
C
v
ideo
coding
st
a
nda
rd
E
l A
n
s
a
ri
Abde
s
s
a
m
a
d
1
,
Nej
m
e
dd
i
ne
B
a
hri
2
,
Ana
s
s
M
a
ns
o
uri
3
,
No
uri
M
a
s
m
o
ud
4
,
Ah
a
it
o
uf
Ali
5
1,
5
F
a
c
u
lt
y
o
f
S
c
ien
c
e
s a
n
d
T
e
c
h
n
o
lo
g
y
,
L
ERS
I,
L
a
b
o
ra
to
ry
,
Un
iv
e
r
sity
o
f
S
id
i
M
o
h
a
m
m
e
d
Be
n
A
b
d
e
ll
a
h
F
e
z
,
M
o
ro
c
c
o
2
,4
Na
ti
o
n
a
l
S
c
h
o
o
l
o
f
En
g
in
e
e
rs
/
L
ET
I
L
a
b
o
ra
to
ry
,
Un
iv
e
rsit
y
o
f
S
f
a
x
,
T
u
n
isia
3
Na
ti
o
n
a
l
S
c
h
o
o
l
o
f
A
p
p
li
e
d
S
c
ien
c
e
s,
L
ERS
I,
L
a
b
o
ra
to
ry
,
Un
iv
e
r
sity
o
f
S
id
i
M
o
h
a
m
m
e
d
Be
n
A
b
d
e
ll
a
h
F
e
z
,
M
o
ro
c
c
o
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
A
p
r
2
2
,
2
0
1
8
R
ev
i
s
ed
Ma
r
8
,
2
0
1
9
A
cc
ep
ted
Ma
r
1
2
,
2
0
1
9
In
th
is
p
a
p
e
r,
w
e
p
ro
p
o
se
a
n
e
w
p
a
ra
ll
e
l
h
a
rd
w
a
re
a
r
c
h
it
e
c
tu
re
f
o
r
th
e
m
o
d
e
d
e
c
isio
n
a
lg
o
rit
h
m
,
th
a
t
it
is
b
a
s
e
d
o
n
th
e
S
u
m
A
b
so
lu
te
o
f
th
e
Diff
e
r
e
n
c
e
(S
A
D)
f
o
r
c
o
m
p
u
te
th
e
m
o
ti
o
n
e
st
im
a
ti
o
n
,
w
h
ich
is
th
e
m
o
st
c
rit
ica
l
a
lg
o
rit
h
m
in
t
h
e
re
c
e
n
t
v
i
d
e
o
e
n
c
o
d
i
n
g
sta
n
d
a
rd
HE
V
C.
In
f
a
c
t,
th
is sta
n
d
a
rd
in
tro
d
u
c
e
d
n
e
w
larg
e
v
a
riab
le
b
lo
c
k
siz
e
o
f
th
e
m
o
ti
o
n
e
stim
a
ti
o
n
a
lg
o
rit
h
m
a
n
d
t
h
e
re
f
o
re
th
e
S
A
D
re
q
u
ire
a
m
o
re
re
d
u
c
e
d
e
x
e
c
u
ti
o
n
t
im
e
in
o
rd
e
r
to
a
c
h
ie
v
e
th
e
re
a
l
ti
m
e
p
ro
c
e
ss
in
g
e
v
e
n
f
o
r
th
e
u
lt
r
a
-
h
ig
h
re
so
l
u
ti
o
n
se
q
u
e
n
c
e
s.
T
h
e
p
r
o
p
o
se
d
a
c
c
e
ler
a
to
r
e
x
e
c
u
tes
th
e
S
A
D
a
lg
o
rit
h
m
in
a
p
a
ra
ll
e
l
w
a
y
f
o
r
a
ll
su
b
-
b
l
o
c
k
p
re
d
ictio
n
u
n
it
s
(
P
Us
)
a
n
d
c
o
d
in
g
u
n
it
(CU)
w
h
a
tev
e
r
th
e
ir
siz
e
s,
w
h
ich
tu
r
n
s
in
a
h
u
g
e
im
p
ro
v
e
m
e
n
ts
in
th
e
p
e
rf
o
r
m
a
n
c
e
s,
g
iv
e
n
th
a
t
a
ll
th
e
b
l
o
c
k
siz
e
s,
P
Us
in
e
a
c
h
CU,
a
re
su
p
p
o
rted
a
n
d
p
ro
c
e
ss
e
d
i
n
t
h
e
sa
m
e
ti
m
e
.
T
h
e
X
il
i
n
x
A
rti
x
-
7
(Z
y
n
q
-
7
0
0
0
)
F
P
G
A
is
u
se
d
f
o
r
th
e
p
ro
t
o
ty
p
in
g
a
n
d
th
e
sy
n
th
e
sis
o
f
th
e
p
r
o
p
o
se
d
a
c
c
e
lera
to
r.
T
h
e
m
o
d
e
d
e
c
isio
n
f
o
r
t
h
e
m
o
ti
o
n
e
stim
a
t
io
n
sc
h
e
m
e
is
im
p
le
m
e
n
ted
w
it
h
3
2
K
L
U
T
s,
5
0
K
r
e
g
isters
a
n
d
1
0
8
Kb
BRA
M
s.
T
h
e
im
p
le
m
e
n
tatio
n
re
su
lt
s
s
h
o
w
th
a
t
o
u
r
h
a
r
d
w
a
re
a
rc
h
it
e
c
tu
re
c
a
n
a
c
h
iev
e
3
0
f
ra
m
e
s
p
e
r
se
c
o
n
d
o
f
th
e
4
K
(3
8
40
×
2
1
6
0
)
re
so
l
u
ti
o
n
s
in
re
a
l
ti
m
e
p
ro
c
e
ss
in
g
a
t
1
1
5
.
1
5
M
Hz
.
K
ey
w
o
r
d
s
:
HE
VC
I
n
teg
er
Mo
tio
n
E
s
ti
m
atio
n
(
I
ME
)
R
ea
l ti
m
e
p
r
o
ce
s
s
in
g
Su
m
A
b
s
o
lu
te
o
f
t
h
e
Dif
f
er
e
n
ce
(
SA
D)
Ver
y
-
L
ar
g
e
-
Scale
I
n
te
g
r
atio
n
(
VL
SI)
ar
ch
itect
u
r
e
Co
p
y
rig
h
t
©
2
0
1
9
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
E
l A
n
s
ar
i
A
b
d
ess
a
m
ad
,
Facu
lt
y
o
f
Sc
ien
ce
s
an
d
T
ec
h
n
o
lo
g
y
o
f
Fez,
L
ab
o
r
ato
r
y
o
f
R
en
e
w
ab
le
E
n
er
g
y
&
S
m
ar
t S
y
s
te
m
s
,
Sid
i M
o
h
a
m
m
ed
B
en
A
b
d
ella
h
Un
iv
er
s
it
y
,
P
.
O.
B
o
x
2
2
0
2
Fez,
0
0
2
1
2
6
2
6
2
4
3
3
5
5
,
M
o
r
o
cc
o
.
E
m
ail:
ab
d
ess
a
m
ad
.
elan
s
ar
i@
u
s
m
b
a.
ac
.
m
a
1.
I
NT
RO
D
UCT
I
O
N
Hig
h
E
f
f
icie
n
c
y
Vid
eo
C
o
d
in
g
(
HE
VC
)
is
th
e
n
e
w
s
tan
d
ar
d
v
id
eo
co
d
in
g
p
r
o
p
o
s
ed
in
J
an
u
ar
y
2
0
1
3
b
y
t
h
e
I
SO/IE
C
an
d
I
T
U
-
T
.
I
t
p
r
o
d
u
ce
s
ef
f
icie
n
c
y
co
d
in
g
u
p
to
5
0
%
co
m
p
ar
ed
to
its
p
r
ed
ec
ess
o
r
H.
2
6
4
/A
VC
[
1
,
2
]
.
T
h
is
n
e
w
s
ta
n
d
ar
d
in
tr
o
d
u
ce
s
i
m
p
r
o
v
e
m
e
n
t
s
in
th
e
e
n
co
d
er
to
w
o
r
k
w
it
h
u
l
tr
a
-
h
i
g
h
r
eso
lu
tio
n
li
k
e
4
k
(
3
8
4
0
×
2
1
6
0
p
ix
els)
an
d
8
k
(
7
6
8
0
×
4
3
2
0
p
i
x
els).
T
h
e
m
ar
k
ed
i
m
p
r
o
v
e
m
e
n
ts
ar
e:
(
i)
th
e
i
n
tr
a
p
r
ed
ictio
n
th
at
w
o
r
k
s
w
it
h
3
5
m
o
d
e
s
i
n
s
tead
o
f
9
m
o
d
es
i
n
H.
2
6
4
;
(
ii)
th
e
s
e
v
en
-
tap
a
n
d
e
ig
h
t
-
tap
f
i
lter
s
u
s
ed
f
o
r
th
e
m
o
tio
n
p
r
ec
is
io
n
,
w
h
e
r
ea
s
o
n
l
y
a
s
i
x
-
tap
i
s
u
s
ed
in
H.
2
6
4
/A
VC
;
a
n
d
(
iii)
th
e
b
lo
c
k
p
ar
titi
o
n
s
tr
u
ct
u
r
e
b
ased
o
n
co
d
in
g
tr
ee
u
n
its
(
C
T
Us),
allo
w
i
n
g
p
r
ed
icti
o
n
r
ef
i
n
e
m
en
t
w
it
h
v
ar
iab
l
e
b
lo
ck
s
ize
f
r
o
m
16
×
1
6
t
o
6
4
×
6
4
ag
ain
s
t th
e
f
i
x
ed
Ma
cr
o
b
lo
ck
s
ize
1
6
×
1
6
i
n
th
e
o
ld
s
ta
n
d
ar
d
H.
2
6
4
/A
VC
.
Ma
n
y
w
o
r
k
s
h
a
v
e
b
ee
n
d
ev
o
ted
to
im
p
le
m
e
n
ti
n
g
t
h
e
HE
V
C
s
tan
d
ar
d
b
y
u
s
i
n
g
b
o
th
s
o
f
t
w
ar
e
an
d
h
ar
d
w
ar
e
f
ac
il
ities
to
r
ea
ch
r
ea
l
ti
m
e
p
r
o
ce
s
s
i
n
g
f
o
r
t
h
e
u
l
tr
a
-
HD
(
h
i
g
h
d
ef
i
n
itio
n
)
ap
p
licatio
n
s
.
T
h
e
v
id
eo
co
d
in
g
al
g
o
r
ith
m
i
s
k
n
o
w
n
b
y
h
i
g
h
co
m
p
u
tin
g
co
m
p
le
x
i
t
y
t
h
at
f
o
r
ce
s
d
esig
n
er
s
to
c
h
o
o
s
e
b
et
w
ee
n
t
h
e
ar
ch
itect
u
r
es
o
f
t
h
e
m
o
r
e
p
o
p
u
lar
Dig
ital
Sig
n
al
P
r
o
ce
s
s
i
n
g
(
DSP
)
an
d
Field
P
r
o
g
r
a
m
m
ab
le
Gate
A
r
r
a
y
(
FP
GA
)
,
w
h
ic
h
ar
e
th
e
m
o
r
e
u
s
ed
e
m
b
ed
d
ed
p
latf
o
r
m
s
.
C
u
r
r
en
tl
y
,
t
h
e
r
ec
en
t
e
m
b
ed
d
ed
s
y
s
te
m
s
o
f
f
er
,
i
n
m
o
s
t
ca
s
es,
b
o
th
h
o
m
o
g
en
o
u
s
an
d
h
eter
o
g
en
eo
u
s
ar
ch
itect
u
r
es
lead
in
g
to
m
a
n
y
i
m
p
le
m
e
n
t
atio
n
p
o
s
s
ib
ilit
ie
s
,
in
cl
u
d
in
g
b
o
th
s
i
m
p
le
an
d
m
i
x
ed
s
o
f
t
w
ar
e
an
d
/o
r
h
ar
d
w
ar
e
i
m
p
le
m
e
n
tat
io
n
s
.
Ho
w
e
v
er
,
ev
en
if
D
SP
s
h
a
v
e
an
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t
201
9
:
2
4
6
9
-
2480
2470
ad
v
an
ta
g
e
i
n
t
h
e
s
p
ee
d
u
p
o
f
h
ig
h
p
r
o
ce
s
s
in
g
al
g
o
r
it
h
m
s
,
t
h
e
y
h
av
e
li
m
itatio
n
s
,
f
o
r
in
s
tan
c
e,
th
e
n
ee
d
ed
ti
m
e
f
o
r
s
o
m
e
s
o
p
h
is
t
icate
d
p
r
o
g
r
am
s
lo
ad
i
n
g
,
w
h
ile
FP
G
A
s
ca
n
ea
s
il
y
s
u
r
r
o
u
n
d
s
u
ch
p
r
o
b
le
m
s
,
e
s
p
ec
iall
y
f
o
r
in
te
n
s
i
v
e
tas
k
s
,
h
av
in
g
m
an
y
l
o
ad
in
g
s
tep
s
.
Ma
n
y
e
f
f
o
r
ts
ar
e
n
ec
e
s
s
ar
y
to
i
d
en
tify
w
h
ic
h
o
f
t
h
e
h
ar
d
w
ar
e
o
r
s
o
f
t
w
ar
e
i
m
p
le
m
e
n
tatio
n
ca
n
b
e
m
o
r
e
ad
v
an
ta
g
eo
u
s
f
o
r
ea
ch
tas
k
.
Fo
r
in
s
ta
n
ce
,
it
is
p
r
ef
er
ab
le
to
co
n
v
er
t
t
h
e
m
o
s
t
ti
m
e
co
n
s
u
m
i
n
g
p
ar
t
in
a
h
ar
d
w
ar
e
p
ar
t,
as
th
e
ca
s
e
o
f
task
s
u
s
in
g
co
m
m
u
n
icatio
n
ex
ch
a
n
g
es
b
et
w
ee
n
b
lo
ck
s
.
So
m
eti
m
e
s
,
m
ix
in
g
t
h
e
ad
v
an
ta
g
e
s
o
f
s
o
f
t
w
ar
e
an
d
h
ar
d
w
ar
e
ca
n
b
e
m
o
r
e
b
en
ef
ic
ial
i
n
t
h
e
w
h
o
l
e
v
id
eo
co
d
in
g
p
r
o
ce
s
s
.
T
h
e
wa
y
is
f
ir
s
t
to
p
r
o
f
ile
th
e
s
o
f
t
w
ar
e
p
r
o
j
ec
t
to
p
in
p
o
in
t
th
e
m
o
r
e
cu
m
b
er
s
o
m
e
p
ar
ts
a
n
d
to
d
ec
id
e
w
h
ic
h
k
in
d
o
f
d
esi
g
n
is
m
o
r
e
ad
ap
ted
.
Fo
r
tu
n
ate
l
y
,
m
an
y
e
m
b
ed
d
ed
d
ev
elo
p
m
en
t
b
o
ar
d
s
o
f
f
er
th
e
m
i
x
in
g
d
esi
g
n
f
ac
ilit
y
a
m
o
n
g
w
h
ic
h
is
th
e
C
h
ip
Z
y
n
q
7
0
0
0
f
r
o
m
th
e
Xili
n
x
f
a
m
il
y
,
b
ased
o
n
2
8
n
m
C
MO
S
t
ec
h
n
o
lo
g
y
.
T
h
e
in
ter
p
r
ed
ictio
n
is
th
e
m
o
r
e
lo
ad
in
g
ti
m
e
b
lo
ck
o
f
th
e
en
co
d
er
,
b
ec
au
s
e
it
co
n
tain
s
th
e
a
lg
o
r
ith
m
s
o
f
m
o
t
io
n
esti
m
a
ti
o
n
an
d
m
o
tio
n
co
m
p
e
n
s
at
io
n
[
3
]
.
T
h
e
m
o
tio
n
esti
m
atio
n
alg
o
r
ith
m
s
t
ill
th
e
s
a
m
e
as
f
o
r
th
e
H.
2
6
4
/A
V
C
,
b
ased
o
n
th
e
co
s
t
S
A
D
w
it
h
th
e
ab
ilit
y
to
p
r
o
ce
s
s
th
e
b
ig
g
e
s
t P
Us s
izes a
n
n
o
u
n
ce
d
in
t
h
e
HE
VC
.
I
n
a
p
r
ev
io
u
s
w
o
r
k
[
4
]
,
w
e
h
av
e
co
m
p
ar
ed
an
i
m
p
le
m
en
ta
tio
n
o
f
a
w
h
o
le
HE
V
C
en
co
d
er
in
b
o
th
p
r
o
ce
s
s
o
r
s
,
AR
M
a
n
d
I
n
tel
in
o
r
d
er
to
s
tu
d
y
t
h
e
s
o
f
t
w
ar
e
co
m
p
lex
i
t
y
.
S
o
,
in
I
n
tel
p
r
o
ce
s
s
o
r
is
f
a
s
ter
ab
o
u
t
1
0
ti
m
e
s
th
an
t
h
e
e
m
b
ed
d
ed
p
r
o
ce
s
s
o
r
A
R
M,
h
o
w
e
v
er
,
th
e
ex
ec
u
t
io
n
s
teel
to
o
f
ar
f
r
o
m
r
ea
l
ti
m
e
p
r
o
ce
s
s
in
g
co
n
d
itio
n
s
,
e
s
p
ec
iall
y
f
o
r
h
ig
h
r
eso
lu
tio
n
.
T
o
o
v
er
co
m
e
th
is
p
r
o
b
lem
,
w
e
f
o
cu
s
o
n
t
h
e
h
ar
d
w
ar
e
i
m
p
le
m
e
n
tat
io
n
in
t
h
i
s
w
o
r
k
.
T
h
e
r
ef
er
e
n
ce
s
o
f
t
w
ar
e
o
f
t
h
e
s
tan
d
ar
d
HE
VC
te
s
t
Mo
d
el
(
HM
)
i
s
av
a
ilab
le
o
n
-
lin
e
[
5
].
T
h
e
T
C
o
m
R
DC
o
s
t
class
,
w
h
i
ch
it
is
u
s
ed
f
o
r
R
D
co
s
t
co
m
p
u
tatio
n
an
d
it
in
cl
u
d
es
t
h
e
all
f
u
n
ctio
n
s
f
o
r
S
A
D
co
m
p
u
tatio
n
.
T
h
e
T
C
o
m
R
DC
o
s
t
tak
es
th
e
lo
n
g
est
e
x
ec
u
t
io
n
ti
m
e
o
f
HM
en
co
d
er
,
w
h
en
co
m
p
ar
ed
to
th
e
o
th
er
class
es.
T
h
at
tak
e
s
ab
o
u
t 4
0
% o
f
th
e
co
m
p
lete
l
y
e
n
co
d
in
g
t
i
m
e
[
3
]
.
I
n
th
e
in
ter
-
p
r
ed
ictio
n
b
lo
ck
s
,
th
e
S
A
D
f
o
r
th
e
m
o
tio
n
es
ti
m
a
tio
n
alg
o
r
it
h
m
i
s
th
e
m
o
s
t
i
m
p
o
r
tan
t
an
d
ti
m
e
-
co
n
s
u
m
i
n
g
s
tep
.
I
ts
h
ar
d
w
ar
e
i
m
p
le
m
en
ta
tio
n
ca
n
b
e
h
elp
f
u
l
f
o
r
t
h
e
v
id
eo
en
co
d
er
.
I
n
th
e
li
ter
atu
r
e,
m
an
y
w
o
r
k
s
ar
e
p
r
o
p
o
s
ed
to
s
u
p
p
o
r
t
th
e
SA
D
ar
ch
itec
tu
r
e
f
o
r
s
ev
e
r
al
ap
p
licatio
n
d
o
m
a
i
n
s
,
s
u
ch
as c
o
m
p
u
ter
v
is
io
n
,
lik
e
m
o
tio
n
d
etec
tio
n
f
o
r
i
m
a
g
e
p
r
o
ce
s
s
i
n
g
[
6
]
,
o
n
th
e
s
y
s
t
e
m
v
id
eo
s
u
r
v
eilla
n
ce
b
ased
m
o
tio
n
d
etec
t
io
n
a
n
d
r
ec
o
g
n
itio
n
in
[
7
]
,
th
at
it
i
m
p
l
e
m
en
t
s
o
n
a
n
e
m
b
ed
d
ed
b
o
ar
d
b
ased
o
n
XC
2
V1
0
0
0
FP
GA
an
d
m
o
tio
n
es
ti
m
atio
n
f
o
r
v
id
eo
co
m
p
r
ess
io
n
s
ta
n
d
ar
d
s
[
8
,
9
,
1
0
,
1
1
,
1
2
,
1
3
,
1
4
]
.
A
ll
th
e
p
r
o
p
o
s
ed
ar
ch
itectu
r
es
aim
to
r
ea
ch
r
ea
l
-
ti
m
e
p
r
o
ce
s
s
in
g
f
o
r
h
i
g
h
er
r
eso
lu
ti
o
n
s
s
eq
u
e
n
ce
s
w
it
h
th
e
h
i
g
h
e
s
t
p
o
s
s
ib
le
o
p
er
atin
g
f
r
eq
u
e
n
c
y
an
d
to
co
m
p
u
te
th
e
m
ax
i
m
u
m
i
n
ter
-
p
r
ed
ictio
n
b
lo
cs
w
h
atev
er
th
eir
s
izes.
W
alter
et
al.
[
8
]
r
ep
o
r
ted
o
n
a
h
ar
d
w
a
r
e
ar
ch
itect
u
r
e
w
it
h
a
h
i
g
h
-
t
h
r
o
u
g
h
p
u
t
f
o
r
lo
w
p
o
w
er
S
A
D
ca
lcu
la
tio
n
f
u
n
ct
i
o
n
in
g
u
p
to
r
ea
l
-
ti
m
e
e
n
co
d
in
g
p
r
o
c
ess
f
o
r
th
e
r
eso
lu
tio
n
o
f
7
2
0
×
4
8
0
p
ix
els.
T
h
eir
d
esig
n
w
as
d
e
v
elo
p
ed
an
d
s
y
n
t
h
e
s
ized
u
s
in
g
t
w
o
C
MO
S
tec
h
n
o
lo
g
ies,
1
8
0
n
m
an
d
6
5
n
m
.
T
h
e
y
u
s
ed
tw
o
m
e
tr
ics
to
co
m
p
ar
e
th
eir
o
b
tain
ed
r
esu
lt
s
,
th
e
m
ax
i
m
u
m
th
r
o
u
g
h
p
u
t
an
d
th
e
m
i
n
i
m
u
m
co
n
s
u
m
ed
e
n
er
g
y
p
er
o
p
er
atio
n
.
T
h
ey
r
ep
o
r
ted
th
e
av
er
a
g
e
o
f
t
h
e
co
n
s
u
m
ed
p
o
w
er
u
s
i
n
g
6
5
n
m
tech
n
o
lo
g
y
d
ec
r
ea
s
es
b
y
5
3
%
co
m
p
ar
ed
t
o
1
8
0
n
m
tec
h
n
o
lo
g
y
,
co
n
s
id
er
in
g
th
e
n
ee
d
ed
lo
w
er
s
u
p
p
l
y
v
o
lta
g
e.
I
n
ad
d
itio
n
,
th
e
n
e
w
tec
h
n
o
l
o
g
y
6
5
n
m
ac
h
ie
v
ed
t
h
e
h
i
g
h
f
r
eq
u
e
n
c
y
b
e
ca
u
s
e
o
f
t
h
e
h
ig
h
er
t
h
r
o
u
g
h
p
u
t.
I
n
[
9
]
,
R
eh
m
a
n
et
al.
p
r
o
p
o
s
ed
ef
f
icie
n
t
h
ar
d
w
ar
e
ar
ch
itec
tu
r
e
f
o
r
th
e
S
AD
alg
o
r
it
h
m
,
d
esi
g
n
ed
f
o
r
i
m
a
g
e
p
r
o
ce
s
s
in
g
.
T
h
e
ad
v
a
n
ta
g
e
o
f
t
h
is
ar
c
h
itect
u
r
e
is
t
h
e
d
ec
r
ea
s
i
n
g
ab
s
o
lu
te
d
i
f
f
er
en
ce
ad
d
itio
n
s
f
o
r
th
e
4
×
4
p
ix
el
s
b
lo
ck
s
izes.
U
s
in
g
a
n
FP
G
A
with
Xil
in
x
X
C
2
V1
0
0
0
,
th
e
y
r
ea
ch
ed
a
f
r
eq
u
e
n
c
y
o
f
1
3
3
MH
z
an
d
u
s
ed
6
5
7
L
UT
s
w
it
h
s
in
g
le
4
×
4
b
lo
c
k
,
an
d
t
h
er
ea
f
ter
,
th
e
y
u
s
e
t
h
e
d
ev
elo
p
ed
4
×
4
en
g
in
e
f
o
r
p
r
o
ce
s
s
in
g
th
e
b
i
g
g
est
b
lo
ck
s
.
An
o
th
er
S
AD
d
ed
icate
d
h
ar
d
w
ar
e
ar
ch
itectu
r
e
h
a
s
b
ee
n
r
ep
o
r
ted
b
y
Kala
m
ir
o
s
an
d
L
y
g
o
u
r
a
s
[
10
]
f
o
r
th
e
s
ter
eo
v
is
io
n
ac
ce
ler
atio
n
,
e
s
p
ec
iall
y
f
o
r
co
m
p
u
ti
n
g
th
e
m
et
h
o
d
o
f
lo
ca
l
co
r
r
elatio
n
.
T
h
e
p
r
o
p
o
s
ed
ar
ch
itectu
r
e
w
a
s
i
m
p
l
e
m
en
ted
o
n
FP
G
A
C
y
clo
n
e
I
I
E
P
2
C
3
5
d
ev
ice,
w
o
r
k
i
n
g
at
1
0
0
MH
z,
an
d
th
eir
p
r
o
p
o
s
ed
SA
D
ac
ce
ler
ato
r
tak
es
2
3
9
0
0
lo
g
ic
ele
m
e
n
t
s
to
r
ea
ch
1
6
2
f
r
a
m
es
p
er
s
ec
o
n
d
in
VG
A
r
e
s
o
lu
tio
n
(
6
4
0
×
4
8
0
)
c
o
d
ed
o
n
8
-
b
it
d
ep
th
.
Z
h
en
y
u
et
al.
[
11
]
d
esig
n
ed
t
wo
h
ar
d
w
ar
e
ar
ch
itect
u
r
es
f
o
r
t
h
e
S
A
D
al
g
o
r
ith
m
,
w
it
h
th
e
H
.
2
6
4
/A
VC
s
tan
d
ar
d
s
u
p
p
o
r
tin
g
t
h
e
v
ar
ia
b
le
b
lo
ck
s
ize
tech
n
iq
u
e.
T
h
e
f
ir
s
t
ar
ch
itec
tu
r
e,
b
ased
o
n
th
e
p
r
o
p
ag
ated
p
ar
tial
S
A
D,
r
ea
ch
es
2
3
1
.
6
MH
z
-
o
p
er
atin
g
f
r
eq
u
e
n
c
y
at
a
co
s
t
o
f
8
4
.
1
K
g
ates,
an
d
th
e
s
ec
o
n
d
i
s
a
tr
ee
b
ased
SA
D
ar
ch
itect
u
r
e
ac
h
ie
v
i
n
g
2
0
4
.
8
MH
z
w
ith
8
8
.
5
k
g
ates.
T
h
e
t
w
o
d
esi
g
n
s
w
er
e
s
y
n
th
e
s
ized
o
n
a
0
.
1
8
µm
1
P
6
M
C
MO
S
tech
n
o
lo
g
y
,
a
n
d
th
e
a
r
ch
itect
u
r
e
lev
e
l
ap
p
r
o
ac
h
es
w
er
e
u
s
ed
to
i
m
p
r
o
v
e
p
er
f
o
r
m
an
ce
.
I
n
ad
d
itio
n
,
P
u
r
n
ac
h
a
n
d
et
al.
[
12
]
d
ev
elo
p
ed
a
h
ar
d
w
ar
e
ar
ch
itect
u
r
e
f
o
r
th
e
HE
VC
v
id
eo
co
d
in
g
s
ta
n
d
ar
d
.
T
h
e
s
y
n
t
h
es
is
w
a
s
p
er
f
o
r
m
ed
o
n
a
Xilin
x
Vir
tex
-
5
FP
GA
,
tak
es
2
9
1
.
2
7
K
o
f
g
ate
co
u
n
ts
,
a
n
d
r
ea
ch
es
1
7
1
.
9
MH
z
as
o
p
er
atin
g
f
r
eq
u
en
c
y
.
T
h
eir
ac
ce
ler
ato
r
is
b
ased
o
n
p
ar
allel
p
r
o
ce
s
s
in
g
a
n
d
d
esig
n
ed
to
s
u
p
p
o
r
t
a
ll
P
Us
b
lo
ck
s
izes,
in
cl
u
d
in
g
b
lo
ck
s
f
o
r
A
s
y
m
m
etr
ic
Mo
tio
n
P
a
r
titi
o
n
in
g
(
AM
P
)
.
Xu
et
al.
[
13
]
an
d
s
y
n
th
e
tized
in
Xilin
x
Vir
tex
-
6
XC
6
V
L
X
-
5
5
0
T
FP
G
A
p
r
o
p
o
s
ed
an
o
th
er
h
ar
d
w
ar
e
S
A
D
ac
ce
ler
ato
r
.
T
h
eir
ar
ch
itec
tu
r
e
co
n
tai
n
s
t
w
o
p
ar
allel
s
u
b
-
ar
ch
itect
u
r
es
to
p
r
o
ce
s
s
1
0
8
0
p
w
i
th
3
0
f
r
a
m
e
s
p
er
s
ec
o
n
d
(
f
p
s
)
r
ea
ch
in
g
r
ea
l
-
ti
m
e
v
id
eo
co
d
in
g
.
Ho
w
e
v
er
,
th
e
ac
ce
ler
ato
r
ca
n
o
n
l
y
co
m
p
u
te
1
6
×
1
6
P
U
an
d
t
h
e
co
r
r
esp
o
n
d
in
g
s
u
b
-
b
lo
ck
s
.
T
h
e
s
y
n
t
h
esi
s
s
h
o
ws
an
o
p
er
atin
g
f
r
eq
u
en
c
y
o
f
1
1
0
MH
z
w
it
h
5
5
3
4
6
L
UT
s
,
1
9
7
4
4
r
eg
is
ter
s
,
a
n
d
1
4
8
k
B
o
f
B
R
A
M.
Mo
r
e
r
ec
en
t
l
y
,
Me
d
h
at
et
al.
[
14
]
p
r
o
p
o
s
ed
a
p
ar
allel
h
ar
d
w
ar
e
S
A
D
ac
ce
le
r
ato
r
f
o
r
th
e
m
o
tio
n
esti
m
a
tio
n
,
s
y
n
th
e
tized
o
n
a
Xilin
x
Vir
ti
x
-
7
XC
7
VX5
5
0
T
FP
GA
.
A
n
o
p
er
atin
g
f
r
eq
u
e
n
c
y
o
f
4
5
8
MH
z
h
a
s
b
ee
n
r
ep
o
r
t
ed
w
it
h
3
9
9
0
1
L
UT
s
an
d
2
4
9
5
7
r
eg
is
ter
s
.
W
ith
th
i
s
ar
ch
itect
u
r
e,
a
2
K
r
eso
lu
tio
n
w
it
h
3
0
f
p
s
ca
n
b
e
p
r
o
ce
s
s
ed
,
w
h
ic
h
is
th
e
s
u
cc
es
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
A
r
ea
and
p
o
w
er e
fficien
t V
LS
I
a
r
ch
itectu
r
e
o
f m
o
d
e
d
ec
is
io
n
i
n
in
teg
er mo
tio
n
…
(
E
l A
n
s
a
r
i A
b
d
ess
a
ma
d
)
2471
k
e
y
o
f
t
h
i
s
ar
ch
i
tectu
r
e.
Ot
h
er
cu
s
to
m
ized
s
o
l
u
tio
n
h
a
v
e
b
ee
n
also
p
u
b
lis
h
ed
s
u
ch
t
h
e
V
L
SI
p
ip
eli
n
ed
m
u
ltip
lier
l
es
s
f
i
x
ed
-
p
o
in
t
8
x
8
DC
T
Desig
n
o
f
a
Fas
t
P
i
p
elin
ed
8
x
8
Dis
cr
ete
C
o
s
in
e
T
r
an
s
f
o
r
m
o
th
er
b
lo
ck
s
[
15
]
.
A
u
th
o
r
s
h
av
e
i
m
p
le
m
en
ted
a
n
d
co
m
p
ar
ed
th
e
DC
T
u
s
in
g
t
w
o
ar
ch
itect
u
r
e
s
,
n
o
n
-
p
ip
elin
e
a
n
d
a
2
-
s
ta
g
e
p
ip
elin
e
a
n
d
th
eir
s
i
m
u
latio
n
r
es
u
lt
s
s
h
o
w
th
a
t
th
r
o
u
g
h
p
u
t
ca
n
b
e
i
m
p
r
o
v
ed
b
y
a
l
m
o
s
t
a
f
ac
to
r
o
f
t
w
o
,
f
r
o
m
ar
o
u
n
d
1
Gig
a
p
ix
els/
s
to
1
.
8
Gig
a
p
ix
els/
s
,
at
a
s
m
a
ll
co
s
t
o
f
r
o
u
g
h
l
y
1
4
%
m
o
r
e
r
eso
u
r
ce
s
(
i.e
.
p
i
p
elin
e
r
eg
is
ter
s
)
.
I
n
th
is
w
o
r
k
,
w
e
p
r
o
p
o
s
e
a
n
e
w
,
h
i
g
h
l
y
e
f
f
icie
n
t
VL
SI
ar
ch
itect
u
r
e
f
o
r
th
e
S
A
D
d
ec
is
i
o
n
m
o
d
e,
i
m
p
le
m
en
ted
in
t
h
e
HM
o
f
HE
VC
.
T
h
is
ar
ch
itectu
r
e
ca
n
p
r
o
ce
s
s
all
s
izes
o
f
P
Us,
in
cl
u
d
in
g
A
MP
,
w
h
ic
h
i
s
u
s
ed
to
i
n
cr
ea
s
e
t
h
e
co
d
in
g
ef
f
icien
c
y
o
f
v
id
eo
s
eq
u
en
ce
s
h
a
v
i
n
g
ar
ea
s
th
a
t
ar
e
m
o
r
e
ir
r
eg
u
lar
.
T
h
e
n
e
w
ar
ch
itect
u
r
e
is
d
esi
g
n
ed
to
lo
ad
all
th
e
i
n
p
u
t
p
ix
el
s
f
r
o
m
t
h
e
r
ef
er
en
ce
a
n
d
cu
r
r
e
n
t
b
lo
ck
s
i
n
a
f
ir
s
t
s
tep
,
m
ak
i
n
g
th
e
m
a
v
ailab
le
d
u
r
i
n
g
t
h
e
p
r
o
ce
s
s
i
n
g
p
r
o
ce
s
s
.
T
h
en
th
e
p
r
o
p
o
s
ed
d
esig
n
h
i
g
h
l
y
r
ed
u
ce
s
ex
ch
a
n
g
es
w
it
h
th
e
m
e
m
o
r
y
a
n
d
th
e
m
e
m
o
r
y
ac
ce
s
s
es
an
d
f
i
n
all
y
lead
s
to
a
p
ar
allel
co
m
p
u
tin
g
o
f
all
P
Us
s
i
m
u
l
tan
eo
u
s
l
y
,
r
ed
u
cin
g
th
e
late
n
c
y
ti
m
e.
2.
H
E
VC
M
O
DE
DE
CIS
I
O
N
O
F
M
O
T
I
O
N
E
ST
I
M
AT
I
O
N
AL
G
O
RI
T
H
M
2
.
1
.
Co
din
g
s
t
ruct
ure
T
h
e
HE
VC
in
tr
o
d
u
ce
s
m
o
r
e
i
m
p
r
o
v
e
m
e
n
t
t
h
an
H.
2
6
4
/
A
VC
[
1
,
2
]
.
T
h
e
m
o
s
t
i
m
p
o
r
tan
ce
is
th
e
h
i
g
h
l
y
f
le
x
ib
le
an
d
ef
f
icie
n
t
b
lo
ck
p
ar
titi
o
n
i
n
g
s
tr
u
ct
u
r
e
b
y
in
tr
o
d
u
cin
g
f
o
u
r
b
lo
ck
co
n
ce
p
ts
:
C
o
d
in
g
T
r
ee
Un
it
(
C
T
U)
,
co
d
in
g
u
n
it
(
C
U)
,
p
r
ed
ictio
n
u
n
it
(
P
U)
,
an
d
tr
a
n
s
f
o
r
m
u
n
it
(
T
U)
.
T
h
e
f
ir
s
t
s
tep
i
s
t
h
e
p
ar
ti
tio
n
in
g
s
tr
u
c
tu
r
e
o
f
th
e
p
ict
u
r
es,
w
h
ic
h
ca
n
b
e
p
ar
titi
o
n
ed
i
n
s
l
ices
a
n
d
tile
s
;
t
h
e
s
lice
is
p
ar
titi
o
n
ed
i
n
to
C
T
Us.
I
n
H.
2
6
4
,
th
e
C
T
Us
ar
e
an
alo
g
u
e
to
a
f
i
x
ed
m
ac
r
o
b
lo
ck
s
ize
o
f
1
6
×1
6
,
w
h
er
ea
s
in
HE
VC
,
t
h
e
y
ca
n
b
e
s
ized
to
1
6
×
1
6
,
3
2
×
3
2
an
d
64
×
6
4
to
ac
ce
p
t
s
ev
er
al
v
id
eo
s
eq
u
en
ce
s
.
T
h
e
s
ize
o
f
a
C
T
U
is
f
i
x
ed
in
th
e
m
ai
n
en
co
d
er
co
n
f
ig
u
r
atio
n
f
ile
an
d
ca
n
b
e
p
ar
titi
o
n
ed
in
to
m
u
ltip
l
e
s
q
u
ar
ed
C
Us r
eg
io
n
s
as
s
h
o
w
n
i
n
F
i
g
u
r
e
1
.
Fig
u
r
e
1
.
C
o
d
in
g
s
tr
u
ct
u
r
e
o
f
t
h
e
C
U
i
n
t
h
e
HE
VC
e
n
co
d
er
T
h
e
n
u
m
b
er
o
f
C
U
s
i
n
a
C
T
U
is
m
ai
n
l
y
d
eter
m
i
n
ed
b
y
th
e
r
eq
u
ir
e
d
r
eso
lu
tio
n
.
T
h
e
s
m
all
C
U
i
s
8
×
8
an
d
ca
n
b
e
s
ca
led
u
p
t
o
6
4
×
6
4
.
T
h
e
C
U
ca
n
b
e
d
iv
id
ed
in
to
P
r
ed
ictio
n
Un
its
.
T
h
is
b
lo
ck
ca
n
b
e
u
s
ed
f
o
r
th
e
in
tr
a
an
d
in
ter
b
lo
ck
p
r
ed
ict
io
n
,
an
d
its
s
ize
v
ar
ies f
r
o
m
4
×
4
to
6
4
×
6
4
.
A
f
ter
th
e
P
U,
an
ap
p
r
o
p
r
iate
s
ize
o
f
th
e
tr
an
s
f
o
r
m
U
n
it,
T
U,
is
d
eter
m
i
n
ed
f
o
r
th
e
I
n
te
g
er
Dir
ec
t Co
s
i
n
e
T
r
an
s
f
o
r
m
(
I
DC
T
)
an
d
th
e
q
u
an
ti
f
icat
io
n
.
I
n
th
e
HE
V
C
en
co
d
er
,
ea
ch
C
U
ca
n
b
e
d
i
v
id
ed
in
to
m
u
ltip
le
P
Us,
an
d
ea
ch
P
U
is
c
h
ar
ac
te
r
ized
b
y
a
m
o
tio
n
es
ti
m
atio
n
v
ec
to
r
.
I
n
ad
d
itio
n
,
w
h
en
t
h
e
C
U
s
ize
is
e
q
u
al
to
2
N×
2
N
w
it
h
N
an
i
n
te
g
e
r
v
al
u
e
a
m
o
n
g
3
2
,
1
6
,
8
o
r
4
d
ep
en
d
in
g
o
n
th
e
d
ep
th
(
0
,
1
,
2
an
d
3
)
in
th
e
co
r
r
e
s
p
o
n
d
in
g
co
d
in
g
tr
ee
s
tr
u
ctu
r
e
,
eig
h
t
P
U
s
p
litt
in
g
t
y
p
es
ar
e
d
ef
i
n
ed
as
illu
s
tr
ate
d
in
F
i
g
u
r
e
2
:
t
w
o
s
q
u
ar
ed
r
eg
io
n
(
P
A
R
T
_
2
N×
2
N
an
d
PAR
T
_
N×
N)
an
d
s
i
x
r
ec
tan
g
u
lar
o
n
e
s
,
t
w
o
s
y
m
m
etr
ic
(
P
A
R
T
_
2
N×
N
an
d
P
A
R
T
_
N×
2
N)
,
an
d
f
o
u
r
as
y
m
m
etr
ic
n
a
m
ed
A
MP
(
P
A
R
T
_
2
N×
n
U,
P
A
R
T
_
2
N×
n
D,
P
A
R
T
_
n
L
×2
N,
a
n
d
P
AR
T
_
n
R
×2
N)
w
h
er
e
U,
D,
L
,
an
d
R
m
ea
n
Up
,
Do
w
n
,
L
e
f
t,
an
d
R
i
g
h
t,
r
esp
ec
tiv
el
y
a
n
d
n
r
ep
r
esen
t
s
th
e
s
m
al
ler
s
iz
e
in
A
MP
p
ar
titi
o
n
m
o
d
es
[
1
6
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t
201
9
:
2
4
6
9
-
2480
2472
Fig
u
r
e
2.
S
y
m
m
e
tr
ic
an
d
as
y
m
m
e
tr
ic
b
lo
ck
p
o
r
tio
n
in
g
o
f
t
h
e
P
U
2
.2
.
M
o
de
decisi
o
n SAD
a
lg
o
rit
h
m
i
n H
E
VC
T
h
e
cr
itical
p
ar
t
o
f
th
e
all
-
v
i
d
eo
en
co
d
er
s
tan
d
ar
d
s
ex
p
lo
itin
g
te
m
p
o
r
al
r
ed
u
n
d
an
c
y
is
t
h
e
Mo
tio
n
E
s
ti
m
a
tio
n
(
ME
)
.
T
h
e
r
ec
en
t
HE
VC
h
as
a
m
o
r
e
co
m
p
lex
m
o
tio
n
esti
m
atio
n
alg
o
r
it
h
m
th
a
n
H.
2
6
4
/A
V
C
,
d
u
e
to
th
e
p
ix
el
d
en
s
it
y
i
n
t
h
e
h
i
g
h
p
ictu
r
e
r
eso
lu
t
io
n
p
r
o
ce
s
s
in
g
(
8
K
an
d
4
K)
an
d
th
e
n
e
w
P
U
b
lo
ck
p
ar
titi
o
n
in
g
.
T
h
e
SA
D
p
r
o
ce
s
s
in
g
is
p
er
f
o
r
m
ed
f
o
r
th
e
in
te
g
er
ME
(
I
M
E
)
,
an
d
it
is
ca
lcu
lated
f
o
r
al
l
s
izes
f
o
r
th
e
P
Us
ca
n
d
id
ates.
I
t
is
co
m
p
u
ted
b
as
ed
o
n
th
e
in
f
o
r
m
atio
n
co
m
i
n
g
f
r
o
m
t
h
e
cu
r
r
en
t
o
r
o
r
ig
in
al
b
l
o
ck
(
B
o
)
o
f
th
e
P
U
an
d
th
e
p
r
ed
icted
o
r
th
e
r
ef
e
r
en
ce
b
lo
ck
(
B
r
)
o
b
tain
ed
b
y
th
e
m
o
tio
n
v
ec
to
r
p
r
ed
icto
r
d
er
iv
atio
n
p
r
o
ce
s
s
.
T
h
e
ca
lcu
latio
n
o
f
t
h
e
S
A
D
i
s
p
er
f
o
r
m
ed
u
s
i
n
g
(
1
)
.
11
(
,
)
(
,
)
KL
ro
ij
S
A
D
B
i
j
B
i
j
(
1
)
W
h
er
e
i
an
d
j
ar
e
th
e
co
r
r
esp
o
n
d
in
g
p
i
x
el
i
n
t
h
e
B
o
an
d
B
r
b
lo
ck
s
.
K
a
n
d
L
ar
e
t
h
e
w
id
t
h
a
n
d
h
ei
g
h
t
o
f
ea
ch
s
ea
r
ch
w
i
n
d
o
w
,
r
esp
e
ctiv
el
y
.
in
d
icate
s
t
h
e
p
ix
el
o
f
t
h
e
r
ef
er
en
ce
b
lo
ck
an
d
B
o
(
i,j
)
in
d
icate
s
th
e
p
ix
e
l
o
f
th
e
o
r
ig
in
al
(
c
u
r
r
en
t)
b
lo
ck
.
T
h
e
co
m
p
u
tat
io
n
co
m
p
le
x
it
y
is
i
n
t
h
e
n
u
m
b
er
o
f
ca
lls
a
n
d
th
e
d
if
f
er
en
t
s
ize
S
A
D
f
o
r
I
ME
o
f
th
e
i
n
ter
p
r
ed
ictio
n
.
Hen
ce
,
a
h
ar
d
w
ar
e
ac
ce
ler
ato
r
o
f
SA
D
i
s
r
eq
u
ir
ed
.
3.
P
RO
P
O
SE
D
SAD
H
AR
DW
ARE ARC
H
I
T
E
C
T
UR
E
T
h
e
SA
D
is
a
co
s
t
f
u
n
ctio
n
u
s
ed
to
c
o
m
p
u
te
d
is
to
r
tio
n
b
et
w
ee
n
t
w
o
b
lo
ck
s
.
Se
v
er
al
ar
ch
it
ec
tu
r
es
f
o
r
th
e
S
A
D
ca
lcu
lat
io
n
w
er
e
p
r
o
p
o
s
ed
in
th
e
liter
atu
r
e.
T
h
ey
ar
e
m
ai
n
l
y
b
ased
o
n
o
n
e
o
f
th
e
f
o
l
lo
w
i
n
g
th
r
ee
k
in
d
s
o
f
d
esig
n
s
,
p
ar
allel,
P
ip
elin
e,
a
n
d
s
eq
u
e
n
tial.
T
h
e
s
eq
u
e
n
tia
l d
esig
n
u
s
es
ac
c
u
m
u
la
to
r
to
s
to
r
e
th
e
r
esu
lts
o
f
th
e
ab
s
o
lu
te
v
a
lu
e
i
n
ea
c
h
clo
ck
c
y
cle.
I
t u
s
es a
m
i
n
i
m
u
m
h
ar
d
w
ar
e
u
s
a
g
e,
b
u
t ta
k
es
m
o
r
e
th
an
o
n
e
clo
ck
c
y
cle
t
o
g
iv
e
th
e
o
u
tp
u
t
r
es
u
lt
s
;
s
u
b
s
eq
u
en
tl
y
,
it
is
i
m
p
o
s
s
ib
le
to
r
ea
ch
u
p
r
ea
l
-
ti
m
e
p
r
o
ce
s
s
i
n
g
d
u
e
to
th
e
h
i
g
h
er
r
eso
lu
tio
n
s
tr
ea
t
m
en
t
in
o
u
r
c
ases
.
T
h
e
p
ip
elin
e
ar
ch
i
tectu
r
e,
as
it
u
s
es
a
s
er
ies
o
f
co
n
n
e
cted
s
tag
e
s
,
tak
e
s
i
n
in
p
u
t
a
v
ec
to
r
o
f
p
i
x
els
in
s
te
ad
o
f
o
n
e
p
i
x
el
i
n
p
u
t
a
n
d
g
e
n
er
ates
s
ev
er
al
r
es
u
lts
o
n
o
n
e
lin
e;
f
o
r
in
s
tan
ce
,
in
[
9
]
,
th
e
au
th
o
r
s
u
s
e
a
n
in
p
u
t
v
ec
to
r
o
f
1
2
8
-
b
it
p
r
esen
t
s
1
6
p
ix
els
8
-
b
it
s
a
m
p
le
o
f
4
×
4
b
lo
ck
s
a
m
p
le.
So
th
e
p
ip
elin
e
ar
ch
itect
u
r
e
is
f
a
s
ter
t
h
an
t
h
e
s
eq
u
e
n
tia
l,
b
u
t
it
is
n
o
t
en
o
u
g
h
to
r
ea
ch
th
e
r
ea
l
ti
m
e
f
o
r
h
ig
h
r
eso
l
u
tio
n
.
I
n
[
9
]
,
th
e
p
r
o
p
o
s
ed
ar
ch
itectu
r
e
is
d
esig
n
ed
in
p
ip
eli
n
ed
s
ta
g
es;
h
o
w
ev
er
,
it
r
ea
ch
e
s
a
r
ate
o
f
3
0
f
p
s
o
n
l
y
f
o
r
s
m
al
l seq
u
e
n
ce
s
w
i
th
7
2
0
×
4
8
0
p
ix
els
w
it
h
th
e
co
n
f
i
g
u
r
atio
n
lev
el
3
o
f
H.
2
6
4
/A
V
C
en
co
d
er
.
T
h
e
p
ar
allel
d
esig
n
is
f
a
s
ter
th
an
b
o
th
s
eq
u
en
tial
an
d
p
ip
elin
e,
b
ec
au
s
e
it
is
co
n
s
tr
u
cted
o
n
s
ev
er
a
l
s
lices,
w
h
ic
h
ar
e
p
r
o
ce
s
s
ed
i
n
d
ep
en
d
en
tl
y
o
f
ea
ch
o
th
er
;
co
n
s
eq
u
e
n
tl
y
t
h
e
ar
ea
co
n
s
u
m
p
tio
n
i
s
i
n
cr
ea
s
ed
.
I
n
ad
d
itio
n
,
it
’
s
p
o
s
s
ib
le
to
m
i
x
th
ese
t
y
p
e
s
;
f
o
r
in
s
ta
n
ce
,
th
e
to
tal
ar
ch
itectu
r
e
w
o
r
k
s
p
ar
allel,
an
d
th
e
s
ilices
d
esig
n
w
o
r
k
in
s
eq
u
e
n
tia
l
o
r
p
ip
elin
ed
m
o
d
e
lik
e
in
r
e
f
.
[
1
1
]
.
Fo
r
th
e
u
s
e
o
f
f
u
ll
y
p
ar
allel
ar
ch
itect
u
r
e,
a
h
ig
h
l
y
o
p
tim
ized
d
esig
n
is
n
ee
d
ed
to
r
ed
u
ce
th
e
co
n
s
u
m
ed
ar
ea
an
d
to
k
ee
p
h
ig
h
er
p
er
f
o
r
m
an
ce
s
.
T
h
e
p
a
r
allelis
m
i
n
th
e
h
ar
d
war
e
ar
ch
itectu
r
e
is
ex
p
lo
r
ed
t
o
in
cr
ea
s
e
th
e
ac
ce
ler
ato
r
th
r
o
u
g
h
p
u
t
to
co
m
p
u
te
al
l
S
AD
o
p
er
atio
n
s
o
n
f
o
u
r
clo
ck
c
y
cle
s
,
b
u
t
t
h
e
n
u
m
b
er
o
f
in
p
u
t
s
a
m
p
le
s
p
i
x
e
ls
i
n
cr
ea
s
es
an
d
t
h
e
n
u
m
b
er
o
f
ad
d
er
s
in
th
e
d
ep
t
h
tr
ee
s
ta
g
e
i
n
cr
ea
s
es.
An
o
p
ti
m
izatio
n
is
n
ec
e
s
s
ar
y
to
r
ed
u
c
e
th
e
co
n
s
u
m
ed
ar
ea
an
d
th
e
d
ela
y
in
r
ea
ch
i
n
g
th
e
r
ea
l ti
m
e
e
n
co
d
in
g
s
y
s
te
m
,
p
ar
ticu
lar
l
y
f
o
r
4
K
ap
p
licatio
n
s
.
T
h
e
SA
D
al
g
o
r
ith
m
i
m
p
le
m
e
n
tatio
n
i
s
ac
h
iev
ed
i
n
th
r
ee
m
ai
n
s
tep
s
;
f
ir
s
t
th
e
ab
s
o
l
u
te
d
if
f
er
e
n
ce
b
et
w
ee
n
t
w
o
p
i
x
els
i
s
ca
lcu
la
ted
b
y
t
h
e
p
r
o
ce
s
s
in
g
u
n
it
(
P
r
U)
d
esig
n
ed
i
n
F
i
g
u
r
e
3
.
Nex
t,
th
e
s
u
m
o
f
th
e
s
e
d
if
f
er
e
n
ce
s
i
s
ca
lcu
la
ted
in
o
n
e
s
tep
f
o
r
f
o
u
r
i
n
p
u
t p
ix
e
ls
u
s
i
n
g
t
h
e
d
eig
n
ed
p
r
o
ce
s
s
in
g
ele
m
en
t (
P
E
)
p
r
esen
ted
in
F
i
g
u
r
e
4
.
Fin
al
l
y
,
f
o
r
th
e
wh
o
le
S
A
D
al
g
o
r
it
h
m
,
t
h
e
P
E
i
s
i
m
p
le
m
e
n
ted
1
0
2
4
ti
m
es
to
allo
w
p
r
o
ce
s
s
i
n
g
o
f
all
b
lo
ck
s
izes
f
r
o
m
4
×
8
&
8
×
4
to
6
4
×
64.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
A
r
ea
and
p
o
w
er e
fficien
t V
LS
I
a
r
ch
itectu
r
e
o
f m
o
d
e
d
ec
is
io
n
i
n
in
teg
er mo
tio
n
…
(
E
l A
n
s
a
r
i A
b
d
ess
a
ma
d
)
2473
3
.
1
.
Arc
hite
ct
ure
o
f
a
bs
o
lute
diff
er
ence
T
h
e
P
r
U,
g
iv
en
t
h
e
ab
s
o
lu
te
d
i
f
f
er
en
ce
(
ab
s
)
,
o
p
er
ates
b
y
co
m
p
ar
i
s
o
n
b
et
w
ee
n
th
e
t
w
o
p
ix
els
co
m
in
g
f
r
o
m
th
e
r
ef
er
en
ce
an
d
cu
r
r
en
t
i
m
a
g
es.
I
f
th
e
d
if
f
er
en
ce
is
p
o
s
itiv
e
o
r
eq
u
al
to
ze
r
o
,
th
e
ab
s
is
d
ir
ec
tly
o
b
tain
ed
;
if
n
o
t,
th
e
t
w
o
co
m
p
l
i
m
e
n
ts
o
f
th
e
n
eg
a
tiv
e
ter
m
ar
e
u
s
ed
to
m
ak
e
t
h
e
r
esu
lt
p
o
s
itiv
e.
T
h
is
o
p
er
atio
n
o
f
p
ix
el
co
m
p
ar
is
o
n
ca
n
b
e
d
escr
ib
ed
b
y
t
h
e
f
o
llo
w
in
g
(
2
)
:
,0
(
)
1
,
0
r
o
r
o
r
o
r
o
B
B
B
B
abs
n
o
t
B
B
B
B
(
2
)
T
h
e
p
r
o
p
o
s
ed
h
ar
d
w
ar
e
ar
ch
i
t
ec
tu
r
e
o
f
P
r
U
is
p
r
ese
n
ted
in
F
ig
u
r
e
3
.
I
t
co
n
s
i
s
ts
o
f
a
n
o
p
tim
al
cir
c
u
it
u
s
i
n
g
o
n
e
m
u
ltip
le
x
er
(
MU
X
)
f
o
r
w
h
ic
h
t
h
e
s
elec
tio
n
i
n
p
u
t
co
r
r
esp
o
n
d
s
to
o
n
e
o
f
t
h
e
t
w
o
s
tate
s
o
f
th
e
co
m
p
ar
ato
r
(
C
o
m
p
)
o
u
tp
u
ts
(
i.e
.
P
o
s
itiv
e
o
r
n
eg
ativ
e
v
al
u
e)
.
Dep
en
d
in
g
o
n
th
e
s
i
g
n
o
f
th
e
co
m
p
ar
is
o
n
,
th
e
MU
X
is
g
iv
e
n
in
ea
ch
ca
s
e
th
e
ab
s
o
lu
te
d
if
f
er
e
n
ce
co
m
i
n
g
f
r
o
m
th
e
o
u
tp
u
t
o
f
t
h
e
s
u
b
tr
ac
to
r
(
SUB
)
.
T
h
e
eq
u
atio
n
n
o
t(
B
r
-
B
o
)
+1
is
i
m
p
le
m
en
ted
u
s
in
g
o
n
e
i
n
v
er
ter
f
o
llo
w
ed
b
y
o
n
e
1
b
it
ad
d
er
.
T
h
e
r
esu
lts
ar
e
al
w
a
y
s
8
b
i
ts
-
co
d
ed
.
T
h
e
r
eg
i
s
ter
s
(
R
)
ar
e
u
s
ed
to
s
to
r
e
th
e
v
alu
e
s
o
f
t
h
e
p
r
o
ce
s
s
ed
p
ix
el
s
.
T
o
d
ec
r
ea
s
e
th
e
h
ar
d
w
ar
e
co
s
t,
t
h
e
d
esig
n
o
f
P
r
U
is
o
p
ti
m
ized
b
y
th
e
o
p
ti
m
iz
atio
n
th
e
co
m
p
ar
ato
r
.
Fig
u
r
e
3
.
P
r
o
ce
s
s
in
g
u
n
it (
P
r
U)
f
o
r
th
e
ca
lcu
latio
n
o
f
t
h
e
ab
s
o
lu
te
d
if
f
er
en
ce
v
al
u
e
T
h
e
m
ai
n
f
u
n
ct
io
n
o
f
C
o
m
p
is
to
co
m
p
ar
e
t
w
o
in
p
u
t
s
i
g
n
als
a
n
d
to
r
etu
r
n
t
h
e
r
es
u
lts
o
n
th
r
e
e
o
u
tp
u
ts
,
ea
ch
co
d
ed
in
1
-
b
it.
T
h
e
C
o
m
p
is
a
8
-
b
it
co
m
p
ar
ato
r
.
T
o
r
e
d
u
ce
t
h
e
u
s
ed
ar
ea
,
it
co
n
tain
s
t
h
e
lo
g
ic
g
a
tes
i
n
t
h
e
1
-
bi
t
co
m
p
ar
ato
r
as
s
h
o
w
n
i
n
F
ig
u
r
e
4
a
n
d
u
s
es
th
e
p
r
i
n
ci
p
al
2
:1
co
m
p
r
es
s
o
r
d
escr
ib
ed
in
r
e
f
[
1
7
]
.
I
n
o
u
r
ar
ch
itect
u
r
e
[
1
8
]
,
th
e
C
o
m
c
o
n
s
tr
u
ct
io
n
i
s
b
ased
o
n
a
tr
ee
o
f
1
-
b
it
co
m
p
ar
ato
r
u
s
i
n
g
th
e
m
eth
o
d
o
f
2
:1
co
m
p
r
es
s
o
r
to
co
m
p
ar
e
t
w
o
A
an
d
B
in
p
u
t
p
i
x
els
co
d
ed
i
n
8
-
b
it
an
d
d
eli
v
er
in
g
t
h
e
r
es
u
lt
s
o
f
t
h
e
co
m
p
ar
i
s
o
n
o
f
th
e
o
u
tp
u
t.
Fig
u
r
e
4
.
1
-
b
it c
o
m
p
ar
ato
r
u
s
i
n
g
i
n
t
h
e
C
o
m
p
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t
201
9
:
2
4
6
9
-
2480
2474
T
h
e
P
E
,
as
s
h
o
w
n
in
F
i
g
u
r
e
5
,
is
co
m
p
o
s
ed
f
r
o
m
f
o
u
r
P
r
U,
s
e
v
en
r
eg
i
s
ter
s
,
an
d
th
r
ee
ad
d
er
s
co
m
p
u
te
th
e
s
u
m
o
f
ab
s
o
l
u
te
d
i
f
f
er
e
n
c
e
o
f
ei
g
h
t
p
ix
e
ls
,
f
o
u
r
f
r
o
m
t
h
e
r
ef
er
e
n
ce
B
l
o
ck
a
n
d
f
o
u
r
f
r
o
m
t
h
e
o
r
i
g
in
a
l.
T
h
e
r
esu
lt is
g
iv
e
n
b
y
t
h
e
o
u
tp
u
t si
g
n
al
S4
x
1
,
w
h
ich
i
s
n
o
w
c
o
d
ed
in
1
0
-
b
its
an
d
o
b
tain
ed
in
o
n
e
clo
ck
c
y
cle.
Fig
u
r
e
5
.
P
r
o
ce
s
s
in
g
ele
m
en
t (
P
E
)
f
o
r
ei
g
h
t p
ix
el
i
n
p
u
t
s
an
d
o
n
e
o
u
tp
u
t
T
h
e
g
lo
b
al
SA
D
ca
lc
u
latio
n
ar
ch
itect
u
r
e
is
p
r
esen
ted
i
n
F
i
g
u
r
e
6
.
I
t c
o
n
tain
s
:
1.
T
h
r
ee
4
K
s
in
g
le
p
o
r
t
R
AM
b
lo
ck
s
,
th
e
t
w
o
f
ir
s
t
(
2
×1
0
2
4
×8
b
its
)
ar
e
d
e
d
icate
d
to
th
e
p
ix
els
o
f
th
e
cu
r
r
en
t
an
d
o
r
ig
in
al
b
lo
ck
s
to
r
ag
e,
an
d
th
e
th
ir
d
is
e
m
p
lo
y
ed
to
s
to
r
e
th
e
r
esu
l
ts
o
f
t
h
e
co
m
p
u
ted
S
A
Ds.
2.
1
0
2
4
P
E
ar
r
an
g
ed
in
a
p
ar
allel
w
a
y
to
s
tar
t
f
u
n
ct
io
n
i
n
g
at
t
h
e
s
a
m
e
ti
m
e,
a
n
d
ca
lcu
la
tin
g
li
n
e
b
y
lin
e
t
h
e
p
ar
tial su
m
o
f
t
h
e
ab
s
o
lu
te
d
i
f
f
er
en
ce
s
(
P
SA
D)
f
o
r
th
e
eig
h
t
s
elec
ted
p
ix
els.
3.
A
th
r
ee
ad
d
er
b
lo
ck
s
to
d
eli
v
er
t
h
e
S
AD
b
y
s
u
m
m
i
n
g
t
h
e
P
SA
D
t
w
o
b
y
t
w
o
u
n
ti
l
t
h
e
f
i
n
al
r
es
u
lt
.
T
h
is
p
ar
t is
n
a
m
ed
co
m
b
i
n
at
io
n
ad
d
itio
n
in
F
ig
u
r
e
5
.
Fig
u
r
e
6
.
T
h
e
g
lo
b
al
ar
ch
itectu
r
e
o
f
th
e
S
A
D
al
g
o
r
ith
m
in
I
ME
f
o
r
HE
VC
en
co
d
er
I
n
th
e
h
ar
d
w
ar
e
ar
ch
itectu
r
e,
t
h
e
P
r
Us
is
o
r
g
an
ized
lin
e
b
y
lin
e
in
s
u
c
h
a
w
a
y
t
h
at,
f
r
o
m
e
ig
h
t
P
r
Us,
t
w
o
s
m
all
P
Us
8
×
4
an
d
4
×
8
ca
n
b
e
co
n
s
tr
u
cted
.
Mo
r
eo
v
er
th
e
p
r
o
p
o
s
ed
a
r
ch
itectu
r
e
is
d
esi
g
n
ed
f
o
r
th
e
lar
g
e
s
t
s
ize
b
lo
ck
6
4
×
6
4
an
d
ca
n
ea
s
i
l
y
b
e
ad
a
p
ted
f
o
r
t
h
e
o
th
er
s
m
allest
s
ize
b
lo
ck
s
w
it
h
ad
eq
u
a
te
m
e
m
o
r
y
s
to
r
ag
e
of
th
e
s
tar
t
in
g
p
ix
el
s
g
i
v
i
n
g
in
F
ig
u
r
e
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
A
r
ea
and
p
o
w
er e
fficien
t V
LS
I
a
r
ch
itectu
r
e
o
f m
o
d
e
d
ec
is
io
n
i
n
in
teg
er mo
tio
n
…
(
E
l A
n
s
a
r
i A
b
d
ess
a
ma
d
)
2475
Fig
u
r
e
7
.
L
o
ca
tio
n
o
f
d
ata
i
n
p
u
t p
ix
el
s
S
u
b
-
C
U
6
4
×
64
3
.2
.
H
i
g
h sp
ee
d a
dd
er
CL
A
T
h
e
SA
D
alg
o
r
it
h
m
is
b
ased
o
n
a
b
asic
ar
it
h
m
etic
o
p
er
atio
n
,
s
in
ce
it
co
m
p
r
i
s
es
d
o
u
b
le
s
u
m
m
atio
n
s
.
T
o
m
a
k
e
it
f
aster
,
s
ev
er
al
p
r
o
p
o
s
ed
ar
ch
itectu
r
es
u
s
e
t
h
e
ca
r
r
y
p
r
o
p
ag
atio
n
f
o
r
th
e
b
in
ar
y
ad
d
itio
n
k
n
o
w
n
b
y
its
h
i
g
h
p
er
f
o
r
m
a
n
ce
s
.
T
h
er
ef
o
r
e,
its
i
m
p
le
m
e
n
tatio
n
is
ad
e
q
u
ate
f
o
r
Ver
y
L
ar
g
e
Scale
I
n
t
eg
r
atio
n
(
V
L
SI)
.
I
n
r
ef
er
en
ce
s
[
1
7
]
an
d
[
1
8
]
,
th
e
au
th
o
r
s
u
s
ed
t
w
o
f
as
ter
p
ar
a
llel
ad
d
er
s
,
R
ip
p
le
C
ar
r
y
A
d
d
er
s
(
R
C
A
)
an
d
C
ar
r
y
L
o
o
k
ah
ea
d
A
d
d
er
s
(
C
L
A
)
,
a
n
d
th
e
y
m
a
k
e
a
co
m
p
ar
is
o
n
b
et
w
ee
n
t
h
e
m
b
ased
o
n
th
e
d
ela
y
an
d
t
h
e
ar
ea
co
n
s
u
m
p
tio
n
.
Fir
s
t,
R
C
A
tak
es
m
o
r
e
d
ela
y
th
a
n
t
h
e
C
L
A
b
ec
au
s
e
it
h
as
a
le
n
g
t
h
y
p
r
o
p
ag
atio
n
(
th
r
o
u
g
h
all
ad
d
er
s
)
;
it
s
tar
ts
f
r
o
m
t
h
e
f
ir
s
t
ad
d
er
an
d
it
f
i
n
is
h
es
at
t
h
e
last
o
n
e.
T
h
e
C
L
A
h
as
n
o
ca
r
r
y
p
r
o
p
ag
atio
n
b
ec
au
s
e
t
h
e
ca
r
r
y
o
u
tp
u
t
o
f
c
u
r
r
en
t
ad
d
er
is
co
n
n
ec
ted
to
th
e
ca
r
r
y
i
n
p
u
t
b
it
o
f
th
e
n
e
x
t
ad
d
er
.
In
th
e
ar
ea
,
th
e
R
C
A
r
eq
u
ir
ed
less
s
u
r
f
ac
e
th
a
n
t
h
e
C
L
A
,
s
i
n
ce
t
h
e
la
s
t
h
a
s
a
s
ca
lar
ar
ch
itect
u
r
e
a
n
d
it
n
ee
d
ed
m
o
r
e
co
m
p
o
n
e
n
t
s
.
Giv
i
n
g
th
at
w
e
ar
e
in
ter
ested
in
a
r
ap
id
s
y
s
t
e
m
,
th
e
C
L
A
i
s
m
o
r
e
ad
o
p
te
d
in
o
u
r
h
ar
d
w
ar
e
ac
ce
ler
ato
r
f
o
r
d
esig
n
i
n
g
al
l
ad
d
er
s
,
ev
en
th
o
u
g
h
it
in
cr
ea
s
e
s
th
e
n
ee
d
ed
ar
ea
,
b
u
t
w
it
h
o
u
t
af
f
ec
ti
n
g
th
e
to
tal
s
u
r
f
ac
e;
m
o
r
eo
v
er
,
it
is
s
i
m
p
le
to
d
esig
n
,
g
i
v
e
n
th
a
t
th
e
C
L
A
ar
ch
itect
u
r
e
u
s
e
s
b
asic
g
ate
l
o
g
ic
s
u
ch
a
s
A
ND
,
OR
an
d
XO
R
.
Ou
r
C
L
A
co
m
p
r
is
e
s
t
w
o
lev
e
l
s
.
T
h
e
f
ir
s
t
lev
el
is
u
s
ed
to
c
o
m
p
u
te
th
e
p
ar
tial
(
in
ter
m
ed
iat
e)
v
alu
es,
Gi
an
d
P
i,
an
d
g
en
er
ate
t
h
e
ter
m
s
,
SUM(
i)
an
d
C
i+1
as
o
u
tp
u
t,
d
ef
i
n
ed
b
y
th
e
eq
u
atio
n
(
3
)
,
(
4
)
,
(
5
)
an
d
(
6
)
as
p
r
esen
ted
in
F
ig
u
r
e
8
,
w
h
ic
h
p
r
o
p
o
s
ed
a
s
ec
tio
n
o
f
th
e
w
h
o
le
C
L
A
d
esi
g
n
.
T
h
e
s
ec
o
n
d
lev
el
is
th
e
to
tal
d
esig
n
o
f
C
L
A
as
s
h
o
w
n
i
n
F
i
g
u
r
e
9
,
w
h
ic
h
co
m
p
u
te
s
ea
ch
ca
r
r
y
,
th
e
to
tal
s
u
m
,
a
n
d
in
d
icate
s
t
h
e
o
v
er
f
lo
w
o
f
ad
d
itio
n
as
o
u
tp
u
t.
W
h
e
n
ea
ch
ca
r
r
y
i
s
av
ailab
le,
th
e
p
ar
tial
s
u
m
is
ca
lcu
lated
s
i
m
u
lta
n
eo
u
s
l
y
a
n
d
th
en
th
e
o
v
er
f
lo
w
o
f
th
e
ad
d
itio
n
is
ca
lc
u
lated
at
t
h
e
en
d
.
G
i
=
A
(
i
)
.B
(
i
)
(
3
)
P
i
=
A
(
i
)
⊕
B
(
i
)
(
4
)
C
o
ut
i
=
G
i
+
(
P
i
.C
i
n
i
)
.
(
5
)
S
u
m
(
i
)
=
P
i
⊕
C
i
(
6
)
Fig
u
r
e
8
.
P
r
o
p
o
s
ed
s
ec
tio
n
o
f
C
L
A
ad
d
er
Fig
u
r
e
9
.
P
r
o
p
o
s
ed
C
L
A
n
-
b
it
ad
d
er
b
ased
o
n
th
e
ar
ch
itect
u
r
e
o
f
C
L
A
ad
d
er
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t
201
9
:
2
4
6
9
-
2480
2476
C
o
u
t_
i
(
o
r
C
n
+1
)
an
d
C
i
n
_
i
(
o
r
C
n
)
co
r
r
esp
o
n
d
r
esp
ec
tiv
el
y
to
th
e
ca
r
r
y
o
u
tp
u
t
a
n
d
th
e
ca
r
r
y
in
p
u
t
o
f
s
ec
tio
n
N.
As
illu
s
tr
ated
in
F
i
g
u
r
e
9
,
ea
ch
C
o
u
t
o
f
a
g
iv
e
n
s
ec
tio
n
is
co
n
n
ec
ted
to
C
in
o
f
t
h
e
s
u
b
s
eq
u
en
t
o
n
e.
Gn
an
d
P
n
ar
e
r
esp
ec
tiv
el
y
t
h
e
g
en
er
ated
an
d
t
h
e
p
r
o
p
ag
ated
ca
r
r
ies.
I
n
Fi
g
u
r
e
9
,
o
u
r
h
ig
h
p
ar
all
el
ar
ch
itect
u
r
e
s
tar
ts
p
r
o
ce
s
s
i
n
g
f
r
o
m
1
0
2
4
P
E
s
an
d
ar
r
iv
es
to
t
h
e
S
A
D_
6
4
×6
4
,
d
if
f
er
en
t
to
th
e
ca
s
e
in
[
1
9
]
,
th
at
ca
n
co
m
p
u
te
t
h
e
s
ize
o
f
w
in
d
o
w
s
d
if
f
er
en
tl
y
.
T
h
e
SA
D_
6
4
×6
4
is
th
e
v
a
lu
e
o
f
S
AD
co
r
r
esp
o
n
d
s
to
th
e
b
ig
g
er
b
o
ck
6
4
×
6
4
,
an
d
g
en
er
ates
all
it
s
co
r
r
esp
o
n
d
in
g
s
m
a
ll
S
A
D
s
i
m
u
lta
n
eo
u
s
l
y
.
T
o
s
u
m
m
ar
iz
e
o
u
r
ap
p
r
o
ac
h
,
o
n
e
c
o
m
p
u
te
s
b
o
t
h
s
m
aller
b
lo
ck
s
8
×
4
an
d
4
×
8
,
p
r
esen
ted
b
y
v
alu
e
s
S
AD_
4
×
8
an
d
S
A
D
_
8
×
4
r
esp
ec
tiv
el
y
.
T
h
e
y
ar
e
co
m
p
u
ted
b
y
t
h
e
ad
d
itio
n
o
f
th
e
ei
g
h
t
P
E
s
co
r
r
es
p
o
n
d
in
g
to
ea
ch
b
lo
ck
.
Af
ter
co
m
p
u
ti
n
g
all
s
m
aller
b
lo
ck
s
in
s
id
e
6
4
×
6
4
,
th
e
n
ex
t
s
tep
co
m
p
u
tes
t
h
e
S
A
D
v
al
u
e
(
SAD_
8
×8
)
o
f
8
×
8
b
lo
ck
b
y
ad
d
in
g
t
w
o
s
m
alle
r
b
lo
ck
s
alr
ea
d
y
p
r
o
ce
s
s
ed
.
F
ol
lo
w
i
n
g
t
h
e
s
a
m
e
m
et
h
o
d
,
th
e
b
lo
ck
is
p
r
o
ce
s
s
e
d
f
r
o
m
t
h
e
p
r
ev
io
u
s
S
A
D
v
al
u
es
to
ar
r
iv
e
at
th
e
v
al
u
e
S
A
D_
64
×
6
4
o
f
th
e
b
ig
g
er
b
lo
ck
.
Fin
all
y
,
th
e
d
esi
g
n
is
ar
r
an
g
ed
in
t
h
e
tr
ee
.
A
n
e
x
a
m
p
l
e
is
g
iv
e
n
in
F
ig
u
r
e
1
0
,
s
h
o
w
i
n
g
S
AD_
8
×8
v
alu
e
is
p
r
o
ce
s
s
ed
b
y
ad
d
in
g
t
w
o
S
A
D_
8
×4
.
Fig
u
r
e
1
0
.
C
o
n
s
tr
u
c
tio
n
o
f
t
h
e
SA
Ds 8
×
4
an
d
4
×
8
f
r
o
m
P
E
s
an
d
also
th
e
co
n
s
tr
u
ctio
n
o
f
8
×
8
f
r
o
m
t
w
o
8
×
4
4.
E
XP
E
R
I
M
E
NT
A
L
RE
SUL
T
S
I
n
th
e
f
ir
s
t step
,
th
e
p
r
o
p
o
s
ed
ar
ch
itect
u
r
e
f
o
r
th
e
S
A
D
al
g
o
r
ith
m
w
as i
m
p
le
m
e
n
ted
in
C
++
lan
g
u
ag
e,
th
en
it w
as
co
m
p
iled
to
b
e
v
alid
ated
an
d
to
r
e
d
u
ce
its
co
m
p
le
x
it
y
.
Af
ter
w
ar
d
s
,
th
e
ac
ce
ler
ato
r
h
ar
d
w
ar
e
d
esig
n
is
d
ev
elo
p
ed
u
s
i
n
g
t
h
e
h
ar
d
war
e
d
escr
ip
tio
n
lan
g
u
a
g
e
VH
DL
,
te
s
ted
u
s
in
g
Me
n
to
r
Gr
ap
h
ics
Mo
d
elSi
m
P
E
b
ef
o
r
e
b
ein
g
s
y
n
th
et
ized
in
t
h
e
FP
GA
A
r
tex
-
7.
T
h
e
f
ir
s
t
te
s
t
w
a
s
p
er
f
o
r
m
ed
o
n
s
o
m
e
te
s
t
v
id
eo
s
eq
u
e
n
ce
s
p
r
o
p
o
s
ed
b
y
t
h
e
s
ta
n
d
ar
d
.
T
h
e
S
A
D
w
a
s
v
alid
ated
u
s
i
n
g
th
r
ee
tech
n
ic
s
,
f
ir
s
t
w
it
h
r
ef
er
en
ce
s
o
f
t
w
ar
e
o
f
HE
VC
HM
1
5
.
0
v
er
s
io
n
[
5
]
,
th
e
s
ec
o
n
d
b
y
t
h
e
s
o
f
t
w
ar
e
C
++
ex
ec
u
ted
u
n
d
er
L
in
u
x
Ub
u
n
t
u
,
an
d
th
e
last
f
r
o
m
th
e
p
r
o
p
o
s
ed
h
ar
d
w
ar
e
d
esig
n
.
T
h
e
th
r
ee
o
b
tain
ed
r
esu
lts
ar
e
i
n
r
ath
er
g
o
o
d
ag
r
ee
m
e
n
t.
T
h
e
f
r
am
e
s
o
f
s
eq
u
e
n
ce
B
ask
e
tb
alld
r
iv
e
(
1
0
8
0
p
)
an
d
o
th
er
s
ar
e
u
s
ed
in
th
is
test
,
an
d
th
e
v
alu
e
o
f
QP
is
f
i
x
ed
at
3
2
.
Fo
r
m
o
r
e
co
n
f
ig
u
r
at
io
n
s
o
f
th
e
ex
ec
u
tio
n
,
w
e
h
a
v
e
b
ee
n
in
s
p
ir
ed
b
y
t
h
e
w
o
r
k
p
r
esen
ted
in
R
ef
.
[
2
0
]
,
w
h
ic
h
p
r
o
v
id
e
a
p
er
f
o
r
m
an
ce
ev
al
u
atio
n
o
f
d
is
tr
i
b
u
ted
HE
VC
en
co
d
in
g
,
a
n
d
w
a
s
f
o
cu
s
ed
o
n
t
h
e
ef
f
ec
ts
o
f
t
h
e
v
id
eo
p
ar
titi
o
n
i
n
g
m
et
h
o
d
o
n
o
v
er
all
d
is
tr
ib
u
ted
en
co
d
in
g
p
er
f
o
r
m
a
n
ce
u
s
in
g
a
p
ar
titi
o
n
i
n
g
s
ch
e
m
e
b
ased
o
n
GOP
in
o
r
d
er
to
m
ea
s
u
r
e
th
e
en
co
d
er
ef
f
icie
n
c
y
.
T
h
en
th
e
R
a
n
d
o
m
-
ac
ce
s
s
o
n
l
y
co
n
f
ig
u
r
atio
n
f
ile
i
s
s
elec
ted
f
r
o
m
t
h
r
ee
co
n
f
i
g
u
r
atio
n
s
(
R
a
n
d
o
m
-
ac
ce
s
s
,
lo
w
-
d
ela
y
a
n
d
in
tr
a
o
n
l
y
)
,
b
ec
au
s
e
it
is
m
o
s
t
s
u
itab
le
f
o
r
o
u
r
ca
s
e.
A
l
l
t
h
ese
c
h
ar
ac
ter
is
tics
a
n
d
e
v
en
m
o
r
e
i
n
f
o
r
m
a
tio
n
ar
e
m
e
n
tio
n
ed
i
n
t
h
e
co
n
f
i
g
u
r
atio
n
f
il
e
p
r
o
p
o
s
ed
b
y
th
e
s
tan
d
ar
d
in
[
2
1
]
.
T
h
e
p
r
o
p
o
s
ed
d
esig
n
b
lo
ck
f
o
r
ar
ch
itectu
r
e
f
o
r
th
e
S
A
D
al
g
o
r
ith
m
i
n
HE
VC
i
s
f
ir
s
t
co
d
ed
in
a
h
ig
h
-
lev
el
lan
g
u
ag
e
C
++
an
d
it
is
i
m
p
le
m
e
n
ted
o
n
b
o
th
p
r
o
ce
s
s
o
r
s
,
A
R
M
co
r
tex
-
a9
e
m
b
ed
d
ed
in
Z
y
n
q
d
ev
ice
an
d
I
n
tel
co
r
e
i3
C
P
U
clo
ck
ed
6
6
7
MH
z
an
d
2
.
4
0
GHz
,
r
esp
ec
tiv
e
l
y
,
to
b
e
v
alid
ated
an
d
t
o
r
e
d
u
ce
th
e
co
m
p
u
tatio
n
a
l
co
m
p
le
x
it
y
.
Seco
n
d
,
th
is
d
esi
g
n
is
al
s
o
co
d
ed
in
VHDL
la
n
g
u
ag
e
a
n
d
test
ed
u
s
i
n
g
Me
n
to
r
Gr
ap
h
ics
Mo
d
elSi
m
P
E
,
th
en
th
e
h
ar
d
w
ar
e
ar
ch
i
tectu
r
e
p
r
o
to
ty
p
ed
o
n
A
r
te
x
-
7
FP
GA
(
x
c7
z0
2
0
-
1
clg
4
0
0
)
f
r
o
m
Xi
lin
x
,
a
n
d
ac
ce
ler
atio
n
h
ar
d
w
ar
e
w
a
s
v
er
if
ied
o
n
Mic
r
o
ze
d
b
o
ar
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
A
r
ea
and
p
o
w
er e
fficien
t V
LS
I
a
r
ch
itectu
r
e
o
f m
o
d
e
d
ec
is
io
n
i
n
in
teg
er mo
tio
n
…
(
E
l A
n
s
a
r
i A
b
d
ess
a
ma
d
)
2477
4
.
1
Sy
nthesis
re
s
ults a
n
d a
na
ly
s
i
s
T
h
e
p
r
o
p
o
s
ed
ar
ch
itectu
r
e
w
a
s
s
y
n
t
h
esized
u
s
in
g
r
ec
en
t
d
es
ig
n
e
n
v
ir
o
n
m
e
n
t
(
Xili
n
x
Viv
a
d
o
2
0
1
6
.
3
)
lik
e
u
s
i
n
g
in
th
e
p
r
ev
io
u
s
w
o
r
k
f
o
r
I
C
T
alg
o
r
ith
m
i
n
s
id
e
t
h
e
HE
V
C
en
co
d
er
[
2
2
]
an
d
in
o
th
er
w
o
r
k
[
2
3
]
.
T
h
e
s
y
n
th
e
s
is
r
esu
lts
f
o
r
o
u
r
ar
ch
itect
u
r
e
ar
e
p
r
esen
ted
i
n
T
ab
le
1
.
T
h
e
d
esig
n
w
o
r
k
s
w
it
h
a
m
a
x
i
m
u
m
f
r
eq
u
en
c
y
u
p
to
1
1
5
.
1
5
MH
z,
an
d
it
co
n
s
u
m
e
s
3
2
7
5
2
L
UT
s
,
co
r
r
esp
o
n
d
in
g
to
6
1
.
5
6
%
o
f
d
ev
ice
r
eso
u
r
ce
s
(
5
3
2
0
0
L
UT
s
)
,
an
d
it
u
s
es
5
0
7
7
r
eg
is
ter
s
(
4
.
7
7
%)
f
r
o
m
t
h
e
1
0
6
4
0
0
av
ailab
le
o
n
th
e
b
o
ar
d
.
I
t
tak
es
1
0
8
Kb
in
th
e
e
m
b
ed
d
ed
m
e
m
o
r
y
B
R
AM
,
an
d
in
ad
d
itio
n
,
o
u
r
d
esig
n
ex
p
lo
its
b
o
th
em
b
ed
d
ed
m
u
ltip
le
x
er
s
o
n
th
e
FP
GA
,
F7
MU
X
an
d
F8
MU
X
b
y
1
5
7
7
an
d
4
8
1
,
r
esp
ec
tiv
el
y
.
Ho
w
e
v
er
,
it
n
ee
d
s
n
o
DSP
s
f
r
o
m
th
e
2
2
0
ex
is
tin
g
DSP
s
in
th
e
FP
GA
.
T
h
e
p
r
o
p
o
s
ed
d
esig
n
w
as
s
i
m
u
lated
-
u
s
i
n
g
Mo
d
elSi
m
.
I
t
w
as
s
h
o
w
n
th
at
th
e
ar
ch
itect
u
r
e
co
u
ld
,
in
o
n
e
clo
ck
c
y
cle,
co
m
p
u
te
th
e
S
A
D
f
o
r
all
b
lo
ck
s
izes
u
p
to
64
×
64
,
in
clu
d
i
n
g
A
MP
.
Giv
in
g
t
h
e
r
ea
ch
ed
f
r
eq
u
e
n
c
y
1
1
5
.
1
5
MH
z,
o
u
r
h
ar
d
w
ar
e
d
e
s
ig
n
ca
n
p
r
o
ce
s
s
3
0
f
r
a
m
es
p
e
r
s
ec
o
n
d
(
f
p
s
)
f
o
r
4
K
r
eso
lu
ti
o
n
ca
lcu
lated
f
r
o
m
th
e
n
u
m
b
er
o
f
s
ea
r
ch
p
o
in
ts
esti
m
ated
f
o
r
3
0
f
p
s
@
4
k
in
s
ec
t
io
n
I
I
d
iv
id
ed
b
y
th
e
o
p
er
atin
g
f
r
eq
u
e
n
c
y
,
a
n
d
it
ca
n
r
ea
ch
u
p
to
1
2
0
f
p
s
f
o
r
th
e
f
u
ll
h
i
g
h
r
eso
l
u
tio
n
ca
lc
u
lated
in
th
e
s
a
m
e
w
a
y
.
T
h
e
d
esig
n
ca
n
b
e
i
m
p
r
o
v
ed
to
r
ea
ch
m
o
r
e
t
h
an
t
h
i
s
n
u
m
b
er
o
f
f
r
a
m
es
p
er
s
ec
o
n
d
w
ith
a
co
s
t
o
f
ad
d
it
io
n
al
r
eso
u
r
ce
co
n
s
u
m
p
tio
n
f
r
o
m
th
e
d
ev
ice.
T
ab
le
1
.
T
h
e
s
y
n
t
h
esi
s
r
es
u
l
ts
o
f
th
e
p
r
o
p
o
s
ed
ar
ch
itectu
r
e
A
v
a
i
l
a
b
l
e
S
A
D
b
l
o
c
k
8
x
4
&
4
x
8
t
o
6
4
×
6
4
S
l
i
c
e
L
U
T
s
3
2
7
5
2
S
l
i
c
e
r
e
g
i
st
e
r
s
5
0
7
7
M
a
x
i
m
u
m
f
r
e
q
u
e
n
c
y
1
1
5
.
1
5
M
h
z
D
S
P
0
F
7
M
U
X
1
5
7
7
F
8
M
U
X
4
8
1
M
e
mo
r
y
1
0
8
K
b
R
A
M
4
.
2
Co
m
pa
riso
n w
it
h o
t
her
wo
r
ks
I
n
T
ab
le
2
,
w
e
r
ep
o
r
t
p
u
b
lis
h
ed
r
esu
lts
[
1
2
,
1
3
,
1
4
]
,
[
2
4
]
an
d
[
25
]
to
co
m
p
ar
e
o
u
r
p
er
f
o
r
m
a
n
ce
s
.
We
s
elec
ted
r
elate
d
w
o
r
k
s
o
n
FP
GA
m
en
t
io
n
i
n
g
t
h
e
s
a
m
e
ch
ar
ac
ter
is
tic
s
,
s
u
c
h
as
s
lice
L
UT
s
,
s
lice
r
eg
is
ter
s
,
m
ax
i
m
u
m
f
r
eq
u
e
n
c
y
,
f
r
a
m
e
r
ate
p
er
s
ec
o
n
d
,
an
d
th
e
P
Us
s
ize
b
lo
ck
u
s
ed
f
o
r
th
e
SA
D
co
m
p
u
tin
g
in
ea
ch
w
o
r
k
.
T
ab
le
2
.
T
h
e
s
y
n
t
h
esi
s
r
es
u
lts
o
f
t
h
e
p
r
o
p
o
s
ed
a
r
ch
itectu
r
e
Our
work
[12
]
[13
]
[14
]
[24
]
[25
]
T
ec
h
n
o
l
o
g
y
2
8
n
m
Ar
t
i
x
-
7
6
5
n
m
Vi
rt
ex
-
5
4
0
n
m
Vi
rt
ex
-
6
2
8
n
m
Vi
rt
ex
-
7
9
0
n
m
Vi
rt
ex
-
4
1
2
0
n
m
Vi
rt
ex
2
L
UT
s
33K
15K
55k
39K
12K
657
R
eg
i
st
ers
5K
20K
19K
24K
7K
-
M
ax
fr
eq
u
en
cy
1
1
5
.
1
5
M
h
z
1
7
1
.
9
4
M
h
z
110Mhz
4
5
8
.
7
M
h
z
32Mhz
133Mhz
S
erac
h
w
i
n
d
o
w
F
rames
rat
e
±
6
4
4
K@30
fp
s
-
-
±
2
4
1
0
8
0
p
@3
0
fps
±
2
0
2
K@30
fp
s
±
1
6
HD
@5
.
3
4
-
1
0
2
4
x
1
0
2
4
@1
2
7
fps
P
Us
AL
L
AL
L
4
×8
t
o
3
2
×32
AL
L
3
2
×
3
2
4
×
4
I
n
[
1
2
]
,
th
e
w
h
o
le
h
ar
d
w
ar
e
ar
ch
itect
u
r
e
f
o
r
S
A
D
a
lg
o
r
it
h
m
is
p
r
esen
ted
,
i
n
o
r
d
er
to
co
m
p
ar
e
o
u
r
p
er
f
o
r
m
a
n
ce
w
it
h
it
s
w
o
r
k
,
w
e
li
m
ited
t
h
e
co
m
p
ar
is
o
n
o
n
l
y
o
n
th
e
co
n
s
u
m
e
ar
ea
o
b
tain
ed
in
t
h
e
Vir
te
x
-
5
6
5
n
m
b
ec
au
s
e
t
h
e
au
t
h
o
r
s
n
o
t
in
d
icate
h
o
w
its
d
esi
g
n
t
h
e
n
u
m
b
er
s
o
f
f
r
a
m
e
ca
n
b
e
p
r
o
ce
s
s
ed
.
s
o
,
its
m
a
x
i
m
u
m
f
r
eq
u
en
c
y
i
s
in
cr
ea
s
ed
ab
o
u
t
1
.
5
ti
m
es
t
h
a
n
o
u
r
f
r
eq
u
e
n
c
y
.
F
o
r
L
UT
s
,
it
tik
es
1
8
K
les
s
t
h
an
o
u
r
w
o
r
k
,
h
o
w
ev
er
,
th
e
r
esu
lts
f
o
r
r
eg
is
ter
s
s
h
o
w
s
th
at
h
a
v
e
1
5
K
m
o
r
e
th
a
n
o
u
r
r
esu
lts
.
Y
u
a
n
et
al.
[
13
]
p
r
esen
t
a
p
ar
allel
S
A
D
ar
ch
itect
u
r
e
s
y
n
t
h
esized
o
n
Vi
r
tex
-
6
4
0
n
m
.
T
h
e
y
u
s
ed
m
o
r
e
L
UT
s
r
eso
u
r
ce
s
a
n
d
r
eg
i
s
ter
s
co
m
p
ar
ed
w
it
h
o
u
r
ca
s
e,
b
u
t
th
e
y
r
ep
o
r
ted
a
lo
w
m
ax
i
m
u
m
f
r
eq
u
en
c
y
,
o
n
l
y
1
1
0
MH
z.
T
h
e
f
r
a
m
es
co
m
p
u
ti
n
g
p
er
s
ec
o
n
d
co
u
ld
b
e
u
p
to
3
0
f
p
s
j
u
s
t
f
u
ll
HD
(
1
0
8
0
p
)
.
Me
d
h
at
el
al.
[
14
]
o
n
an
FP
GA
Vir
tex
-
7
w
it
h
2
8
n
m
,
t
h
e
s
a
m
e
tec
h
n
o
lo
g
y
u
s
ed
in
th
i
s
w
o
r
k
,
h
a
s
r
ep
o
r
ted
an
o
th
er
p
ar
allel
d
esig
n
.
E
v
e
n
if
th
eir
r
esu
lt
s
s
ee
m
co
m
p
ar
ab
le
to
o
u
r
s
,
o
u
r
d
esig
n
tik
es
1
1
8
%
o
f
L
UT
s
less
th
a
n
its
w
o
r
k
,
h
o
w
ev
er
,
o
u
r
r
es
u
lt
o
f
th
e
r
eg
i
s
ter
s
co
n
s
u
m
i
n
g
s
h
o
w
s
th
a
t
h
a
v
e
3
8
0
%
m
o
r
e
t
h
an
o
u
r
r
es
u
lts
w
i
th
a
l
o
w
er
m
a
x
i
m
u
m
f
r
eq
u
en
c
y
(
3
9
8
%
m
o
r
e
th
a
n
o
u
r
f
r
eq
u
e
n
c
y
)
.
T
h
ey
e
x
p
lo
it
f
i
v
e
ti
m
e
s
m
o
r
e
s
lice
r
eg
i
s
ter
s
,
an
d
th
e
y
ca
n
p
r
o
ce
s
s
3
0
f
p
s
f
o
r
th
e
2
K
r
eso
lu
tio
n
,
u
s
in
g
on
l
y
a
s
m
all
s
ea
r
ch
w
in
d
o
w,
104
×
1
0
4
p
ix
els.
C
o
m
p
ar
ed
t
o
o
u
r
ar
ch
itect
u
r
e,
Mu
r
alid
h
ar
a
n
d
R
a
m
ar
o
[
2
4
]
r
ep
o
r
t
o
n
a
lo
w
er
co
n
s
u
m
ed
a
r
ea
(
L
UT
an
d
r
eg
is
ter
)
o
n
th
e
Vir
tex
-
4
FP
GA
,
b
u
t
th
e
y
s
ta
y
ed
s
o
f
ar
f
r
o
m
th
e
r
ea
l
ti
m
e
p
r
o
ce
s
s
i
n
g
,
w
it
h
a
r
ate
j
u
s
t
5
.
3
2
f
p
s
f
o
r
th
e
HD
r
eso
lu
tio
n
;
th
eir
cir
cu
it
is
s
ti
ll
o
n
l
y
w
e
ll
ad
o
p
ted
f
o
r
th
e
H.
2
6
4
/A
VC
v
id
eo
co
d
in
g
s
tan
d
ar
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t
201
9
:
2
4
6
9
-
2480
2478
C
o
m
p
ar
ed
to
ex
is
ti
n
g
w
o
r
k
s
,
o
u
r
d
esig
n
p
r
esen
t
s
a
g
o
o
d
p
er
f
o
r
m
an
ce
t
h
at
e
x
p
lo
r
es
th
e
p
ar
allelis
m
i
n
t
h
e
h
ar
d
w
ar
e
ar
ch
i
tectu
r
e,
w
h
ich
i
s
th
e
w
a
y
to
r
ea
ch
h
i
g
h
th
r
o
u
g
h
p
u
t
f
o
r
4
K
en
co
d
in
g
i
n
r
ea
l
-
t
i
m
e.
T
h
er
ef
o
r
e
,
o
u
r
cir
cu
it
p
r
esen
ts
a
g
o
o
d
co
d
in
g
an
d
ef
f
icie
n
c
y
r
eg
ar
d
in
g
th
e
r
eso
u
r
ce
co
n
s
u
m
p
tio
n
.
J
o
s
h
i
et
al.
[
2
5
]
p
r
esen
ted
an
ar
ch
itect
u
r
e
f
o
r
S
A
D
in
HE
VC
s
ta
n
d
ar
d
.
it
u
s
e
s
p
ip
elin
ed
d
esig
n
w
it
h
p
ar
allel
p
r
o
ce
s
s
in
g
w
it
h
1
6
p
r
o
ce
s
s
in
g
u
n
i
t
in
o
r
d
er
to
r
ed
u
ce
co
m
p
u
t
atio
n
ti
m
e
o
f
th
e
e
x
ec
u
tio
n
a
n
d
to
in
cr
ea
s
e
th
e
p
er
f
o
r
m
a
n
ce
.
T
h
e
h
ar
d
w
ar
e
d
esig
n
is
p
r
o
to
ty
p
ed
an
d
s
i
m
u
lated
o
n
Xili
n
x
Vir
te
x
-
5
FP
G
A
f
o
r
XC
5
VL
X2
0
T
f
a
m
il
y
.
an
d
also
is
s
y
n
t
h
esized
o
n
6
5
n
m
tec
h
n
o
lo
g
y
a
n
d
r
ea
ch
es
a
m
ax
i
m
u
m
clo
ck
f
r
eq
u
en
c
y
o
f
4
7
5
.
2
1
MH
z.
T
h
e
y
d
id
n
o
t
g
i
v
e
a
th
e
n
u
m
b
er
o
f
co
d
ed
f
r
a
m
es,
a
n
d
t
h
e
h
ig
h
es
t
f
r
eq
u
en
c
y
o
b
tai
n
ed
led
to
a
m
o
r
e
co
n
s
u
m
p
tio
n
o
f
en
er
g
y
.
T
h
ey
o
n
l
y
r
ea
c
h
ed
j
u
s
t
1
2
7
f
p
s
f
o
r
1
0
2
4
×
1
0
2
4
im
ag
e
r
eso
lu
tio
n
,
b
u
t t
h
eir
d
esi
g
n
i
s
s
till
li
m
ited
to
th
e
m
id
d
le
r
eso
lu
tio
n
.
5.
CO
NCLU
SI
O
N
I
n
th
i
s
p
ap
er
,
w
e
d
esig
n
ed
a
n
o
v
el
h
ig
h
p
ar
allel
ac
ce
le
r
ato
r
f
o
r
th
e
S
A
D
alg
o
r
it
h
m
h
ar
d
w
ar
e
i
m
p
le
m
en
ta
tio
n
.
T
h
is
co
n
ce
r
n
s
th
e
n
e
w
HE
VC
s
ta
n
d
ar
d
,
an
d
it
is
im
p
le
m
en
ted
o
n
an
FP
GA
A
r
ti
x
-
7
Xilin
x
Z
y
n
q
-
7
0
0
0
.
T
h
e
d
esig
n
is
p
ar
allel
an
d
allo
w
s
t
h
e
ac
c
eler
atio
n
o
f
in
te
g
er
m
o
tio
n
esti
m
atio
n
w
it
h
n
o
m
o
d
i
f
icat
io
n
in
t
h
e
m
o
tio
n
esti
m
atio
n
al
g
o
r
ith
m
.
T
h
e
p
r
o
p
o
s
ed
ar
ch
itectu
r
e
h
as
a
m
ax
i
m
u
m
o
p
er
atin
g
f
r
eq
u
en
c
y
u
p
to
1
1
5
.
1
5
MH
z
an
d
ca
n
p
r
o
ce
s
s
t
h
e
w
ell
-
k
n
o
w
n
4
K
v
id
eo
s
eq
u
e
n
ce
s
with
h
ig
h
r
eso
l
u
tio
n
3840
×
2
1
6
0
w
it
h
i
n
t
h
e
r
ea
l
-
ti
m
e
r
eq
u
ir
e
m
e
n
ts
.
C
o
m
p
ar
ed
w
it
h
p
r
e
v
io
u
s
w
o
r
k
s
,
th
is
d
es
ig
n
p
r
esen
t
s
a
h
i
g
h
s
p
ee
d
an
d
lo
w
ar
ea
co
s
t
ar
c
h
itectu
r
e
s
u
itab
le
f
o
r
v
id
eo
en
co
d
er
s
o
v
er
a
w
i
d
e
r
an
g
e
o
f
v
id
eo
ap
p
licatio
n
s
.
T
h
e
d
esig
n
is
s
i
m
p
le
an
d
p
er
f
e
ctl
y
ad
ap
ted
f
o
r
th
e
HE
VC
en
co
d
er
.
I
n
th
e
n
ex
t w
o
r
k
,
o
u
r
p
r
o
p
o
s
ed
ar
ch
itectu
r
e
w
il
l
b
e
u
s
ed
as
an
e
n
g
in
e
f
o
r
th
e
n
ex
t
p
r
o
p
o
s
ed
cir
cu
it
f
o
r
an
i
n
ter
p
r
ed
ictio
n
b
lo
ck
o
f
a
HE
VC
en
co
d
er
,
k
ee
p
in
g
th
e
r
ea
l
-
t
i
m
e
p
r
o
ce
s
s
i
n
g
f
o
r
h
i
g
h
er
r
eso
l
u
tio
n
s
,
8
K
an
d
4
K.
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
is
w
o
r
k
is
p
ar
tiall
y
s
u
p
p
o
r
ted
b
y
th
e
Mo
r
o
cc
an
-
T
u
n
i
s
ia
n
p
r
o
g
r
am
n
0
1
7
/T
M
2
4
titl
ed
Stu
d
y
a
n
d
d
esig
n
o
f
a
n
e
m
b
ed
d
ed
s
y
s
t
e
m
f
o
r
Ultr
a
Hi
g
h
De
f
i
n
itio
n
co
m
p
r
es
s
io
n
o
n
a
d
ed
icate
d
m
u
lti
-
co
m
p
o
n
e
n
t
ar
ch
itect
u
r
e,
as
A
p
p
licatio
n
:
Dig
ita
l
T
V.
T
h
e
au
t
h
o
r
s
wo
u
ld
lik
e
to
t
h
a
n
k
t
h
e
Mo
r
o
cc
an
an
d
T
u
n
i
s
ia
n
g
o
v
er
n
m
e
n
ts
f
o
r
th
eir
f
in
a
n
cia
l su
p
p
o
r
t.
RE
F
E
R
E
NC
E
S
[1
]
J.
-
R.
Oh
m
,
G
.
J.
S
u
ll
iv
a
n
,
H
.
S
c
h
w
a
r
z
,
T
.
K.
T
a
n
,
a
n
d
T
.
W
ieg
a
n
d
,
“
Co
m
p
a
riso
n
o
f
th
e
c
o
d
i
n
g
e
ff
ic
ien
c
y
o
f
v
id
e
o
c
o
d
in
g
sta
n
d
a
rd
si
n
c
lu
d
in
g
h
ig
h
e
ff
icie
n
c
y
v
id
e
o
c
o
d
i
n
g
(HE
V
C),
”
Ci
rc
u
it
s
S
y
st.
Vi
d
e
o
T
e
c
h
n
o
l.
IE
EE
T
ra
n
s
.
O
n
,
v
o
l.
2
2
,
n
o
.
1
2
,
p
p
.
1
6
6
9
1
6
8
4
,
2
0
1
2
.
[2
]
G
.
J.
S
u
ll
iv
a
n
,
J.
-
R.
Oh
m
,
W
.
-
J.
Ha
n
,
a
n
d
T
.
W
ieg
a
n
d
,
“
Ov
e
rv
i
e
w
o
f
th
e
h
ig
h
e
ff
icie
n
c
y
v
id
e
o
c
o
d
i
n
g
(HEV
C
)
sta
n
d
a
rd
,
”
Circ
u
i
ts S
y
st.
V
id
e
o
T
e
c
h
n
o
l
.
IEE
E
T
ra
n
s.
On
,
v
o
l.
2
2
,
n
o
.
1
2
,
p
p
.
1
6
4
9
1
6
6
8
,
2
0
1
2
.
[3
]
F
.
B
o
ss
e
n
,
B.
Bro
ss
,
K.
S
u
h
rin
g
,
a
n
d
D.
F
ly
n
n
,
“
HE
V
C
c
o
m
p
lex
it
y
a
n
d
im
p
le
m
e
n
tatio
n
a
n
a
ly
sis,
”
Circ
u
it
s
S
y
st.
Vi
d
e
o
T
e
c
h
n
o
l.
IE
EE
T
r
a
n
s.
O
n
,
v
o
l.
2
2
,
n
o
.
1
2
,
p
p
.
1
6
8
5
1
6
9
6
,
2
0
1
2
.
[4
]
El
A
n
sa
ri,
A
b
d
e
ss
a
m
a
d
,
A
li
A
h
a
i
to
u
f
,
a
n
d
A
n
a
ss
M
a
n
so
u
ri.
“
A
n
im
p
le
m
e
n
tatio
n
c
o
m
p
a
riso
n
o
f
th
e
HEV
C
e
n
c
o
d
e
r
o
n
tw
o
e
m
b
e
d
d
e
d
p
r
o
c
e
ss
o
rs
,
”
W
ire
les
s
T
e
c
h
n
o
lo
g
ies
,
Em
b
e
d
d
e
d
a
n
d
I
n
telli
g
e
n
t
S
y
ste
ms
(
W
IT
S
),
2
0
1
7
I
n
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
IEE
E
,
2
0
1
7
.
[5
]
sv
n
HEV
CS
o
f
tw
a
re
-
Re
v
isio
n
4
5
9
8
:
/
t
a
g
s/HM
-
1
5
.
0
.
[
On
li
n
e
]
.
A
v
a
il
a
b
le:
,
<
h
tt
p
s://
h
e
v
c
.
h
h
i
.
f
ra
u
n
h
o
f
e
r.
d
e
/sv
n
/sv
n
_
HEV
CS
o
f
tw
a
r
e
/t
a
g
s/HM
-
1
5
.
0
/>
[6
]
J.
-
H.
Hs
ieh
,
J.
-
H.
Hu
a
n
g
,
a
n
d
H.
-
R.
W
a
n
g
,
“
DV
F
S
-
a
w
a
re
m
o
ti
o
n
e
stim
a
ti
o
n
d
e
sig
n
sc
h
e
m
e
b
a
s
e
d
o
n
b
a
n
d
w
id
th
ra
ted
isto
rti
o
n
o
p
ti
m
iza
t
io
n
in
a
p
p
li
c
a
ti
o
n
p
ro
c
e
ss
o
r
sy
ste
m
s,
”
In
teg
r.
VL
S
I
J
.
,
v
o
l.
5
7
,
p
p
.
7
4
8
0
,
2
0
1
7
.
[7
]
M
o
u
ss
a
,
M
o
u
ra
d
,
Ne
srin
e
Bd
i
o
u
i
,
a
n
d
A
li
Do
u
ik
,
“
M
o
ti
o
n
De
tec
ti
o
n
a
n
d
Cl
u
ste
rin
g
Us
in
g
P
CA
a
n
d
NN
in
Co
lo
r
Im
a
g
e
S
e
q
u
e
n
c
e
,
”
T
e
lec
o
mm
u
n
ica
ti
o
n
C
o
mp
u
ti
n
g
El
e
c
tro
n
ics
a
n
d
Co
n
tro
l
(
T
EL
KOM
NIKA
),
1
6
.
2
,
7
4
7
-
7
5
4
,
2
0
1
8
.
[8
]
F
.
L
.
W
a
lt
e
r,
C.
M
.
Din
iz,
a
n
d
S
.
Ba
m
p
i,
“
S
y
n
th
e
sis
a
n
d
c
o
m
p
a
riso
n
o
f
lo
w
-
p
o
w
e
r
h
ig
h
-
th
r
o
u
g
h
p
u
t
a
rc
h
it
e
c
tu
re
s
f
o
r
S
A
D ca
l
c
u
latio
n
,
A
n
a
lo
g
In
teg
r,
”
Circ
u
it
s S
i
g
n
a
l
Pro
c
e
ss
.
,
v
o
l.
7
3
,
n
o
.
3
,
p
p
.
8
7
3
8
8
4
,
2
0
1
2
.
[9
]
S
.
Re
h
m
a
n
,
R.
Yo
u
n
g
,
C.
Ch
a
tw
in
,
a
n
d
P
.
Birch
,
“
A
n
F
P
G
A
b
a
se
d
g
e
n
e
ric
f
ra
m
e
w
o
rk
f
o
r
h
ig
h
sp
e
e
d
s
u
m
o
f
a
b
so
lu
te
d
if
fe
re
n
c
e
i
m
p
le
m
e
n
tatio
n
,
”
Eu
r.
J
.
S
c
i.
Res
.
,
v
o
l.
3
3
,
n
o
.
1
,
p
.
6
2
9
,
2
0
0
9
.
[1
0
]
J.
Ka
lo
m
iro
s
a
n
d
J.
Ly
g
o
u
ra
s,
“
Co
m
p
a
ra
ti
v
e
stu
d
y
o
f
lo
c
a
l
S
AD
a
n
d
d
y
n
a
m
ic
p
ro
g
ra
m
m
in
g
f
o
r
s
tere
o
p
ro
c
e
ss
in
g
u
sin
g
d
e
d
ica
ted
h
a
rd
w
a
re
,
”
EURA
S
IP
J
.
Ad
v
.
S
ig
n
a
l
Pro
c
e
ss
.
,
v
o
l.
2
0
0
9
,
n
o
.
1
,
p
p
.
1
1
8
,
2
0
1
0
.
[1
1
]
Z.
L
iu
,
S
.
G
o
to
,
a
n
d
T
.
Ik
e
n
a
g
a
,
“
Op
ti
m
iza
ti
o
n
o
f
P
ro
p
a
g
a
te
P
a
rti
a
l
S
A
D
a
n
d
S
A
D
tree
m
o
ti
o
n
e
stim
a
ti
o
n
h
a
rd
w
ired
e
n
g
in
e
f
o
r
H.
2
6
4
,
”
i
n
C
o
mp
u
ter
De
sig
n
,
2
0
0
8
.
ICCD 2
0
0
8
.
IEE
E
In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
,
p
p
.
3
2
8
3
3
3
,
2
0
0
8
.
[1
2
]
P
.
Na
ll
u
ri
,
L
.
N.
A
lv
e
s,
a
n
d
A
.
Na
v
a
rro
,
“
A
n
o
v
e
l
S
A
D
a
rc
h
it
e
c
tu
re
f
o
r
v
a
riab
le
b
l
o
c
k
siz
e
m
o
ti
o
n
e
stim
a
ti
o
n
i
n
HEV
C
v
id
e
o
c
o
d
in
g
,
i
n
S
y
ste
m
o
n
Ch
i
p
(S
o
C),
”
2
0
1
3
I
n
ter
n
a
ti
o
n
a
l
S
y
mp
o
si
u
m
,
p
p
.
1
4
,
2
0
1
3
.
[1
3
]
X
.
Yu
a
n
,
L
.
Jin
so
n
g
,
G
.
L
i
w
e
i,
Z
.
Zh
i,
a
n
d
R.
K.
T
e
n
g
,
“
A
h
ig
h
p
e
rfo
rm
a
n
c
e
VL
S
I
a
rc
h
it
e
c
tu
re
fo
r
in
teg
e
r
mo
ti
o
n
e
stima
ti
o
n
in
HEVC,
”
in
AS
IC
(
A
S
ICON),
2
0
1
3
IEE
E
1
0
t
h
In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
,
p
p
.
1
4
,
2
0
1
3
.
Evaluation Warning : The document was created with Spire.PDF for Python.