
We present a new fixed mesh algorithm for solving a class of interface
inverse problems for the typical elliptic interface problems. These interface
inverse problems are formulated as shape optimization prob lems whose
objective functionals depend on the shape of the interface. Regardless of the
location of the interface, both the governing partial differential equations
and the objective functional are discretized optimally, with respect to the
involved polynomial space, by an immersed finite element (IFE) method on a
fixed mesh. Furthermore, the formula for the gradient of the descritized
objective function is de rived within the IFE framework that can be computed
accurately and efficiently through the discretized adjoint procedure. Features
of this proposed IFE method based on a fixed mesh are demonstrated by its
applications to three representative interface inverse problems: the interface
inverse problem with an internal measurement on a subdomain, a
DirichletNeumann type inverse problem whose data is given on the boundary, and
a heat dissipation design problem.

In this paper, we propose an interpretable LSTM recurrent neural network,
i.e., multivariable LSTM for time series with exogenous variables. Currently,
widely used attention mechanism in recurrent neural networks mostly focuses on
the temporal aspect of data and falls short of characterizing variable
importance. To this end, our multivariable LSTM equipped with tensorized
hidden states is developed to learn variable specific representations, which
give rise to both temporal and variable level attention. Preliminary
experiments demonstrate comparable prediction performance of multivariable
LSTM w.r.t. encoderdecoder based baselines. More interestingly, variable
importance in real datasets characterized by the variable attention is highly
in line with that determined by statistical Granger causality test, which
exhibits the prospect of multivariable LSTM as a simple and uniform endtoend
framework for both forecasting and knowledge discovery.

DNNs are ubiquitous datacenter workloads, requiring orders of magnitude more
computing power from servers than traditional workloads. As such, datacenter
operators are forced to adopt domainspecific accelerators that employ
halfprecision floatingpoint (FP) numeric representations to improve
arithmetic density. Unfortunately, even these representations are not dense
enough, and are, therefore, suboptimal for DNNs. We propose a hybrid approach
that employs dense block floatingpoint (BFP) arithmetic on dot product
computations and FP arithmetic elsewhere. While using BFP improves the
performance of dot product operations, that compose most of DNN computations,
allowing values to freely float between dot product operations leads to a
better choice of tensor exponents when converting values to back BFP. We show
that models trained with hybrid BFPFP arithmetic either match or outperform
their FP32 counterparts, leading to more compact models and denser arithmetic
in computing platforms.

We construct and analyze a group of immersed finite element (IFE) spaces
formed by linear, bilinear and rotated Q1 polynomials for solving planar
elasticity equation involving interface. The shape functions in these IFE
spaces are constructed through a group of approximate jump conditions such that
the unisolvence of the bilinear and rotated Q1 IFE shape functions are always
guaranteed regardless of the Lam\`e parameters and the interface location. The
boundedness property and a group of identities of the proposed IFE shape
functions are established. A multipoint Taylor expansion is utilized to show
the optimal approximation capabilities for the proposed IFE spaces through the
Lagrange type interpolation operators.

We study the weak Galerkin finite element method for Stokes problem. A new
weak Galerkin finite element velocitypressure space pair is presented which
satisfies the discrete infsup condition. Based on this space pair, we
establish a stable weak Galerkin approximation scheme without adding any
stability term or penalty term. Then, we further derive the optimal error
estimates for velocity and pressure approximations, respectively. Numerical
experiments are provided to illustrate the theoretical analysis.

An event display software SERENA has been designed for the Jiangmen
Underground Neutrino Observatory (JUNO). The software has been developed in the
JUNO offline software system and is based on the ROOT display package EVE. It
provides an essential tool to display detector and event data for better
understanding of the processes in the detectors. The software has been widely
used in JUNO detector optimization, simulation, reconstruction and physics
study.

We present a unified framework for developing and analysing immersed finite
element (IFE) spaces for solving typical elliptic interface problems with
interface independent meshes. This framework allows us to construct a group of
new IFE spaces with either linear, or bilinear, or the rotatedQ1 polynomials.
Functions in these IFE spaces are locally piecewise polynomials defined
according to the subelements formed by the interface itself instead of its
line approximation. We show that the unisolvence for these IFE spaces follows
from the invertibility of the ShermanMorrison matrix. A group of estimates and
identities are established for the interface geometry and shape functions that
are applicable to all of these IFE spaces. Most importantly, these fundamental
preparations enable us to develop a unified multipoint Taylor expansion
procedure for proving that these IFE spaces have the expected optimal
approximation capability according to the involved polynomials.

The Jiangmen Underground Neutrino Observatory (JUNO) is a neutrino experiment
to determine neutrino mass hierarchy. It has a central detector used for
neutrino detection, which consists of a spherical acrylic vessel containing 20
kt LS and about 18,000 20inch PMTs to collect light from LS. Around the CD,
there is a water pool to shield radioactivities. The WP is equipped with about
2000 PMTs to measure cosmic ray muons by detecting Cherenkov light.
As one of the important parts in JUNO offline software, the serial simulation
framework is developed based on SNiPER. It is in charge of physics generator,
detector simulation, event mixing and digitization. However Geant4 based
detector simulation of such a large detector is timeconsuming and challenging.
It is necessary to take full advantages of parallel computing to speedup
simulation. Starting from version 10, Geant4 supports eventlevel parallelism.
Even though based on pthread, it could be extended with other libraries such as
Intel TBB. Therefore it is possible to parallelize JUNO simulation framework
via integrating Geant4 and SNiPER.
In this paper, our progress in developing parallelized simulation software
are presented. The SNiPER framework can run in sequential mode, Intel TBB mode
or other modes. The SNiPER task component is in charge of event loop, which is
like a simplified application manager. Two types of tasks are introduced in the
simulation framework, one is global task and another is worker task. The global
task will run only once to initialize detector geometry and physics processes.
The worker tasks will be spawned after global task is done. In each worker
task, a Geant4 run manager is invoked to do the real simulation. Therefore the
simulation framework and the underlying TBB have been decoupled. Finally, the
software performance of parallelized JUNO simulation software is also
presented.

The Jiangmen Underground Neutrino Observatory (JUNO) detector is designed to
determine the neutrino mass hierarchy and precisely measure oscillation
parameters. The general purpose design also allows measurements of neutrinos
from many terrestrial and nonterrestrial sources. The JUNO Event Data Model
(EDM) plays a central role in the offline software system, it describes the
event data entities through all processing stages for both simulated and
collected data, and provides persistency via the input/output system. Also, the
EDM is designed to enable flexible event handling such as event navigation, as
well as the splitting of MC IBD signals and mixing of MC backgrounds. This
paper describes the design, implementation and performance of JUNO EDM.

JUNO is a multipurpose neutrino experiment which is designed to determine
neutrino mass hierarchy and precisely measure oscillation parameters. As one of
the important systems, the JUNO offline software is being developed using the
SNiPER software. In this proceeding, we focus on the requirements of JUNO
simulation and present the working solution based on the SNiPER.
The JUNO simulation framework is in charge of managing event data, detector
geometries and materials, physics processes, simulation truth information etc.
It glues physics generator, detector simulation and electronics simulation
modules together to achieve a full simulation chain. In the implementation of
the framework, many attractive characteristics of the SNiPER have been used,
such as dynamic loading, flexible flow control, multiple event management and
Python binding. Furthermore, additional efforts have been made to make both
detector and electronics simulation flexible enough to accommodate and optimize
different detector designs.
For the Geant4based detector simulation, each subdetector component is
implemented as a SNiPER tool which is a dynamically loadable and configurable
plugin. So it is possible to select the detector configuration at runtime. The
framework provides the event loop to drive the detector simulation and
interacts with the Geant4 which is implemented as a passive service. All levels
of user actions are wrapped into different customizable tools, so that user
functions can be easily extended by just adding new tools. The electronics
simulation has been implemented by following an event driven scheme. The SNiPER
task component is used to simulate data processing steps in the electronics
modules. The electronics and trigger are synchronized by triggered events
containing possible physics signals.

We take the JUNO experiment as an example to explore nuclearites in the
future large liquid scintillator detector. Comparing to the previous
calculations, the visible energy of nuclearites across the liquid scintillator
will be reestimated for the liquid scintillator based detector. Then the JUNO
sensitivities to the nuclearite flux are presented. It is found that the JUNO
projected sensitivities can be better than $7.7 \times 10^{17} {\rm cm^{2}
s^{1} sr^{1}}$ for the nuclearite mass $10^{15} \; {\rm GeV} \leq M \leq
10^{24}$ GeV and initial velocity $10^{4} \leq \beta_0 \leq 10^{1}$ with a 20
year running. Note that the JUNO will give the most stringent limits for
downgoing nuclearites with $1.6 \times 10^{13} \; {\rm GeV} \leq M \leq 4.0
\times 10^{15}$ GeV and a typical galactic velocity $\beta_0 = 10^{3}$.

Content Centric Networking (CCN) is a new network infrastructure around
content dissemination and retrieval, shift from host addresses to named data.
Each CCN router has a cache to store the chunks passed by it. Therefore the
caching strategy about chunk placement can greatly affect the whole CCN
performance. This paper proposes an implicit coordinate chunk caching location
and searching scheme (CLS) in CCN hierarchical infrastructure. In CLS, there is
at most one copy of a chunk cached on the path between a server and a leaf
router. This copy is pulled down one level towards the leaf router by a request
or pushed up one level towards the server by the cache eviction. Thus, it is
possible to store more diverse contents in the whole CCN and improve the
network performance. Plus, in order to reduce the server workload and file
download time, a caching trail of chunk is created to direct the following
request where to find the chunk. Extensive testbed experiments have been
performed to evaluate the proposed scheme in terms of a wide range of
performance metrics. The results show that the proposed scheme outperforms
existing algorithms.

In this paper, we use a unified framework introduced in [3] to study two
classes of nonconforming immersed finite element (IFE) spaces with integral
value degrees of freedom. The shape functions on interface elements are
piecewise polynomials defined on subelements separated either by the actual
interface or its line approximation. In this unified framework, we use the
invertibility of the well known ShermanMorison systems to prove the existence
and uniqueness of shape functions on each interface element in either
rectangular or triangular mesh. Furthermore, we develop a multiedge expansion
for piecewise functions and a group of identities for nonconforming IFE
functions which enable us to show that these IFE spaces have the optimal
approximation capability.

The Jiangmen Underground Neutrino Observatory (JUNO) is a multipurpose
neutrino experiment designed to measure the neutrino mass hierarchy using a
central detector (CD), which contains 20 kton liquid scintillator (LS)
surrounded by about 17,000 photomultiplier tubes (PMTs). Due to the large
fiducial volume and huge number of PMTs, the simulation of a muon particle
passing through the CD with the Geant4 toolkit becomes an extremely
computationintensive task. This paper presents a fast simulation
implementation using a socalled voxel method: for scintillation photons
generated in a certain LS voxel, the PMT's response is produced beforehand with
Geant4 and then introduced into the simulation at runtime. This
parameterisation method successfully speeds up the most CPU consuming process,
the optical photon's propagation in the LS, by a factor of 50. In the paper,
the comparison of physics performance between fast and full simulation is also
given.

The Jiangmen Underground Neutrino Observatory (JUNO), a 20 kton multipurpose
underground liquid scintillator detector, was proposed with the determination
of the neutrino mass hierarchy as a primary physics goal. It is also capable of
observing neutrinos from terrestrial and extraterrestrial sources, including
supernova burst neutrinos, diffuse supernova neutrino background, geoneutrinos,
atmospheric neutrinos, solar neutrinos, as well as exotic searches such as
nucleon decays, dark matter, sterile neutrinos, etc. We present the physics
motivations and the anticipated performance of the JUNO detector for various
proposed measurements. By detecting reactor antineutrinos from two power plants
at 53km distance, JUNO will determine the neutrino mass hierarchy at a 34
sigma significance with six years of running. The measurement of antineutrino
spectrum will also lead to the precise determination of three out of the six
oscillation parameters to an accuracy of better than 1\%. Neutrino burst from a
typical corecollapse supernova at 10 kpc would lead to ~5000
inversebetadecay events and ~2000 allflavor neutrinoproton elastic
scattering events in JUNO. Detection of DSNB would provide valuable information
on the cosmic starformation rate and the average corecollapsed neutrino
energy spectrum. Geoneutrinos can be detected in JUNO with a rate of ~400
events per year, significantly improving the statistics of existing geoneutrino
samples. The JUNO detector is sensitive to several exotic searches, e.g. proton
decay via the $p\to K^++\bar\nu$ decay channel. The JUNO detector will provide
a unique facility to address many outstanding crucial questions in particle and
astrophysics. It holds the great potential for further advancing our quest to
understanding the fundamental properties of neutrinos, one of the building
blocks of our Universe.

A new immersed finite element (IFE) method is developed for secondorder
elliptic problems with discontinuous diffusion coefficient. The IFE space is
constructed based on the rotated Q1 nonconforming finite elements with the
integralvalue degrees of freedom. The standard nonconforming Galerkin method
is employed in this IFE method without any penalty stabilization term. Error
estimates in energy and L2 norms are proved to be better than $O(h\sqrt{\log
h})$ and $O(h^2\log h)$, respectively, where the logarithm factors reflect
jump discontinuity. Numerical results are reported to confirm our analysis.

The Jiangmen Underground Neutrino Observatory (JUNO) is an experiment
proposed to determine the neutrino mass hierarchy and probe the fundamental
properties of neutrino oscillation. The JUNO central detector is a spherical
liquid scintillator detector with 20 kton fiducial mass. It is required to
achieve a $3\%/\sqrt{E(MeV)}$ energy resolution with very low radioactive
background, which is a big challenge to the detector design. In order to ensure
the detector performance can meet the physics requirements, reliable detector
simulation is necessary to provide useful information for detector design. A
simulation study of natural radioactivity backgrounds in the JUNO central
detector has been performed to guide the detector design and set requirements
to the radiopurity of detector materials.

We present partially penalized immersed finite element methods for solving
parabolic interface problems on Cartesian meshes. Typical semidiscrete and
fully discrete schemes are discussed. Error estimates in an energy norm are
derived. Numerical examples are provided to support theoretical analysis.

In this paper, we derive a priori error estimates for a class of interior
penalty discontinuous Galerkin (DG) methods using immersed finite element (IFE)
functions for a classic secondorder elliptic interface problem. The error
estimation shows that these methods can converge optimally in a meshdependent
energy norm. The combination of IFEs and DG formulation in these methods allows
local mesh refinement in the Cartesian mesh structure for interface problems.
Numerical results are provided to demonstrate the convergence and local mesh
refinement features of these DGIFE methods.

This article presents new immersed finite element (IFE) methods for solving
the popular second order elliptic interface problems on structured Cartesian
meshes even if the involved interfaces have nontrivial geometries. These IFE
methods contain extra stabilization terms introduced only at interface edges
for penalizing the discontinuity in IFE functions. With the enhanced stability
due to the added penalty, not only these IFE methods can be proven to have the
optimal convergence rate in the H1norm provided that the exact solution has
sufficient regularity, but also numerical results indicate that their
convergence rates in both the H1norm and the L2norm do not deteriorate when
the mesh becomes finer which is a shortcoming of the classic IFE methods in
some situations. Trace inequalities are established for both linear and
bilinear IFE functions that are not only critical for the error analysis of
these new IFE methods, but also are of a great potential to be useful in error
analysis for other IFE methods.

Low power design has become one of the most significant requirements when
CMOS technology entered the nanometer era. Therefore, timing budget is often
performed to slow down as many components as possible so that timing slacks can
be applied to reduce the power consumption while maintaining the performance of
the whole design. Retiming is a procedure that involves the relocation of
flipflops (FFs) across logic gates to achieve faster clocking speed. In this
paper we show that the retiming and slack budgeting problem can be formulated
to a convex cost dual network flow problem. Both the theoretical analysis and
experimental results show the efficiency of our approach which can not only
reduce power consumption by 8.9%, but also speedup previous work by 500 times.

As a thin layer of palladium (Pd) is directly deposited on an yttrium iron
garnet or YIG (Y3Fe5O12) magnetic insulator film, Pd develops both low and
highfield magnetotransport effects that are absent in standalone Pd or thick
Pd on YIG. While the lowfield magnetoresistance peak of Pd tracks the coercive
field of the YIG film, the much larger highfield magnetoresistance and the
Hall effect do not show any obvious relationship with the bulk YIG
magnetization. The distinct highfield magnetotransport effects in Pd are
shown to be caused by interfacial local moments in Pd.

We demonstrate the epitaxial growth of EuO on GaAs by reactive molecular beam
epitaxy. Thin films are grown in an adsorptioncontrolled regime with the aid
of an MgO diffusion barrier. Despite the large lattice mismatch, it is shown
that EuO grows well on MgO(001) with excellent magnetic properties. Epitaxy on
GaAs is cubeoncube and longitudinal magnetooptic Kerr effect measurements
demonstrate a large Kerr rotation of 0.57{\deg}, a significant remanent
magnetization, and a Curie temperature of 69 K.

The carrier type and density in Bi2Se3 single crystals are systematically
tuned by introducing a calcium (Ca) dopant. A carrier density of ~1x1017 cm3
which corresponds to ~25 meV in the Fermi energy is obtained in both n and
ptype materials. Electrical transport properties show that the insulating
behavior is achieved in low carrier density crystals. In addition, both the
band gap and reduced effective mass of carriers are determined.

We present an infrared magnetooptical study of the highly thermoelectric
narrowgap semiconductor Bi$_2$Se$_3$. Farinfrared and midinfrared (IR)
reflectance and transmission measurements have been performed in magnetic
fields oriented both parallel and perpendicular to the trigonal $c$ axis of
this layered material, and supplemented with UVvisible ellipsometry to obtain
the optical conductivity $\sigma_1(\omega)$. With lowering of temperature we
observe narrowing of the Drude conductivity due to reduced quasiparticle
scattering, as well as the increase in the absorption edge due to direct
electronic transitions. Magnetic fields $H \parallel c$ dramatically
renormalize and asymmetrically broaden the strongest farIR optical phonon,
indicating interaction of the phonon with the continuum freecarrier spectrum
and significant magnetoelectric coupling. For the perpendicular field
orientation, electronic absorption is enhanced, and the plasma edge is slightly
shifted to higher energies. In both cases the direct transition energy is
softened in magnetic field.