Physical Review Research,
Journal Year:
2023,
Volume and Issue:
5(2)
Published: June 27, 2023
Dynamical
mean-field
theory
(DMFT)
maps
the
local
Green's
function
of
Hubbard
model
to
that
Anderson
impurity
and
thus
gives
an
approximate
solution
from
a
simpler
quantum
model.
Accurate
solutions
nonetheless
become
intractable
for
large
systems.
Quantum
hybrid
quantum-classical
algorithms
have
been
proposed
efficiently
solve
models
by
preparing
evolving
ground
state
under
Hamiltonian
on
computer
is
assumed
scalability
accuracy
far
beyond
current
state-of-the-art
hardware.
As
proof
principle
demonstration
targeting
we,
first
time,
close
DMFT
loop
with
noisy
With
highly
optimized
fast-forwarding
circuit
noise-resilient
spectral
analysis
we
observe
both
metallic
Mott-insulating
phases.
Based
Cartan
decomposition,
our
algorithm
fixed
depth,
fast-forwarding,
can
evolve
initial
over
arbitrarily
long
times
without
time-discretization
errors
typical
other
product
decomposition
formulas
such
as
Trotter
decomposition.
By
exploiting
structure
circuits
reduce
gate
count
(to
77
cnots
after
optimization),
simulate
dynamics,
extract
frequencies
We
then
demonstrate
Mott
transition
mapping
phases
metal-insulator
phase
diagram.
Near
transition,
method
maintains
where
error
would
otherwise
dominate
due
long-time
evolution
required
resolve
quasiparticle
resonance
frequency
extremely
zero.
This
work
presents
computation
sides
using
digital
hardware,
made
viable
in
terms
simulation
error,
runtime
To
inform
future
computations
analyze
versus
time
domain.
Both
algebraic
decompositions
mitigation
techniques
adopted
could
be
applied
attempt
correlated
electronic
phenomena
computers.
We
present
and
implement
a
parquet
approximation
within
the
dual-fermion
formalism
based
on
partial
bosonization
of
dual
vertex
function
which
substantially
reduces
computational
cost
calculation.
The
method
relies
splitting
exactly
into
single-boson
exchange
contributions
residual
four-fermion
vertex,
physically
embody
respectively
long-range
short-range
spatial
correlations.
After
recasting
equations
in
terms
these
are
solved
using
truncated
unity
Eckhardt
et
al.
[Phys.
Rev.
B
101,
155104
(2020)],
allows
for
rapid
convergence
with
number
form
factors
different
regimes.
While
our
numerical
treatment
can
be
restricted
to
only
few
Matsubara
frequencies,
reminiscent
Astretsov
075109
one-
two-particle
spectral
information
is
fully
retained.
In
applications
two-dimensional
Hubbard
model
agrees
quantitatively
stochastic
summation
diagrams
over
wide
range
parameters.
High-temperature
bad-metal
transport
has
been
recently
studied
both
theoretically
and
in
experiments
as
one
of
the
key
signatures
strong
electronic
correlations.
Here
we
use
dynamical
mean
field
theory
its
cluster
extensions,
well
finite-temperature
Lanczos
method
to
explore
influence
lattice
frustration
on
thermodynamic
properties
Hubbard
model
at
high
temperatures.
We
consider
triangular
square
lattices
half-filling
15%
hole
doping.
find
that
for
$T\ensuremath{\gtrsim}1.5t$
self-energy
becomes
practically
local,
while
finite-size
effects
become
small
size
$4\ifmmode\times\else\texttimes\fi{}4$
types
doping
levels.
The
vertex
corrections
optical
conductivity,
which
are
significant
even
temperatures,
contribute
less
lattice.
approximately
linear
temperature
dependence
dc
resistivity
doped
Mott
insulator
lattices.
Carbon Trends,
Journal Year:
2022,
Volume and Issue:
9, P. 100231 - 100231
Published: Oct. 1, 2022
Numerical
approaches
to
the
correlated
electron
problem
have
achieved
considerable
success,
yet
are
still
constrained
by
several
bottlenecks,
including
high
order
polynomial
or
exponential
scaling
in
system
size,
long
autocorrelation
times,
challenges
recognizing
novel
phases,
and
Fermion
sign
problem.
Methods
machine
learning
(ML),
artificial
intelligence,
data
science
promise
help
address
these
limitations
open
up
a
new
frontier
strongly
quantum
simulations.
In
this
paper,
we
review
some
of
progress
area.
We
begin
examining
context
classical
models,
where
their
underpinnings
application
can
be
easily
illustrated
benchmarked.
then
discuss
cases
ML
methods
enabled
scientific
discovery.
Finally,
will
examine
applications
accelerating
model
solutions
state-of-the-art
many-body
like
Monte
Carlo
potential
future
research
directions.
High-${T}_{c}$
cuprate
strange
metals
are
characterized
by
a
DC
resistivity
that
scales
linearly
with
$T$
from
the
onset
of
superconductivity
to
crystal
melting
temperature,
current
life
time
${\ensuremath{\tau}}_{\ensuremath{\hbar}}\ensuremath{\simeq}\ensuremath{\hbar}/({k}_{B}T)$,
``Planckian
dissipation''.
At
same
time,
optical
conductivity
ceases
be
Drude
form
at
high
temperatures,
suggesting
change
underlying
dynamics
surprisingly
leaves
$T$-linear
unaffected.
We
use
AdS/CFT
correspondence
describes
strongly
coupled,
densely
many-body
entangled
metallic
states
matter
study
thermoelectrical
transport
properties
and
conductivities
local
quantum
critical
Gubser-Rocha
holographic
metal
in
$2+1$
dimensions
presence
lattice
potential,
prime
candidate
compare
experiment.
find
electrical
is
linear
low
temperatures
for
large
range
potential
strengths
wave
vectors,
even
as
it
transitions
between
different
dissipative
regimes.
weak
evolves
function
increasing
temperature
``bad
metal''
mid-IR
resonance
without
changing
transport,
similar
seen
metals.
This
peak
notably
its
evolution
can
fully
understood
consequence
umklapp
hydrodynamics:
i.e.,
hydrodynamic
perturbations
Bloch
modes
lattice.
strong
an
``incoherent
realized
instead
where
momentum
conservation
no
longer
plays
role
transport.
confirm
this
regime
thermal
diffusivity
appears
insensitive
breaking
translations
explained
Planckian
dissipation
originating
universal
microscopic
chaos.
A
behavior
has
been
found
homogeneous
relaxation.
The
charge
does
not
submit
chaos
explanation,
though
continuing
linear-in-$T$
saturates
apparent
slope,
numerically
equal
rate.
Physical Review Research,
Journal Year:
2023,
Volume and Issue:
5(2)
Published: June 27, 2023
Dynamical
mean-field
theory
(DMFT)
maps
the
local
Green's
function
of
Hubbard
model
to
that
Anderson
impurity
and
thus
gives
an
approximate
solution
from
a
simpler
quantum
model.
Accurate
solutions
nonetheless
become
intractable
for
large
systems.
Quantum
hybrid
quantum-classical
algorithms
have
been
proposed
efficiently
solve
models
by
preparing
evolving
ground
state
under
Hamiltonian
on
computer
is
assumed
scalability
accuracy
far
beyond
current
state-of-the-art
hardware.
As
proof
principle
demonstration
targeting
we,
first
time,
close
DMFT
loop
with
noisy
With
highly
optimized
fast-forwarding
circuit
noise-resilient
spectral
analysis
we
observe
both
metallic
Mott-insulating
phases.
Based
Cartan
decomposition,
our
algorithm
fixed
depth,
fast-forwarding,
can
evolve
initial
over
arbitrarily
long
times
without
time-discretization
errors
typical
other
product
decomposition
formulas
such
as
Trotter
decomposition.
By
exploiting
structure
circuits
reduce
gate
count
(to
77
cnots
after
optimization),
simulate
dynamics,
extract
frequencies
We
then
demonstrate
Mott
transition
mapping
phases
metal-insulator
phase
diagram.
Near
transition,
method
maintains
where
error
would
otherwise
dominate
due
long-time
evolution
required
resolve
quasiparticle
resonance
frequency
extremely
zero.
This
work
presents
computation
sides
using
digital
hardware,
made
viable
in
terms
simulation
error,
runtime
To
inform
future
computations
analyze
versus
time
domain.
Both
algebraic
decompositions
mitigation
techniques
adopted
could
be
applied
attempt
correlated
electronic
phenomena
computers.