Publications
Parallel Primitives for Domain Decomposition in Neural Networks
We develop a suite of parallel primitives for high-order tensors in distributed deep learning.
Physical characteristics of the stream – lake transitional zone affect littoral lake metabolism
In this study, we used high-frequency dissolved oxygen data and inverse modeling to estimate daily rates of GPP and R in an …
L-Sweeps: A scalable, parallel preconditioner for the high-frequency Helmholtz equation
We present the first fast solver for the high-frequency Helmholtz equation that scales optimally in parallel, for a single right-hand …
DistDL: Distributed Deep Learning for PyTorch
DistDL is a Python package for model parallel deep learning in PyTorch.
A Linear Algebraic Approach to Model Parallelism in Deep Learning
We present a linear algebraic model for distributed deep learning.
A Survey of Computational Tools in Solar Physics
The SunPy Project developed a 13-question survey to understand the software and hardware usage of the solar-physics community. Of the …
The SunPy Project: Open Source Development and Status of the Version 1.0 Core Package
The goal of the SunPy project is to facilitate and promote the use and development of community-led, free, and open source data …
L-Sweeps Experimental Models and Configurations
Public release of model and configuration data for the 2D experiments in the L-Sweeps paper.
The authors thank Total E&P Research …
L-Sweeps 2D: Scalable Parallel Helmholtz Solver
Public release of code for the 2D experiments in the L-Sweeps paper.
The authors thank Total E&P Research and Technology USA for …
The method of polarized traces for the 3D Helmholtz equation
We present a faster solver for the 3D Helmholtz equation at high frequencies, including derivations of the preconditioner, and analysis …
Pollution-free and fast hybridizable discontinuous Galerkin solvers for the high-frequency Helmholtz equation
In this work we propose a hybridizable discontinuous Galerkin (hdG) discretization of the high-frequency Helmholtz equation in the …
A short note on a pipelined polarized-trace algorithm for 3D Helmholtz
We present a fast solver for the 3D high-frequency Helmholtz equation in heterogeneous, constant density, acoustic media. The solver is …
Optimized finite difference coefficients for the Helmholtz equation
Optimized finite difference (OFD) coefficients are often used to minimize numerical dispersion and to improve accuracy in finite …
Preconditioning the 2D Helmholtz equation with polarized traces
We present a domain decomposition solver for the 2D Helmholtz equation, with a special choice of integral transmission condition that …
PySIT: Python Seismic Imaging Toolbox
The Python Seismic Imaging Toolbox (PySIT) is research-scale platform for developing and prototyping numerical methods and algorithms …
Time-stepping beyond CFL: A locally one-dimensional scheme for acoustic wave propagation
In this abstract, we present a case study in the application of a time-stepping method, unconstrained by the CFL condition, for …
High-dimensional wave atoms and compression of seismic data sets
Wave atoms are a low-redundancy alternative to curvelets, suitable for high-dimensional seismic data processing. This abstract extends …
Sunpy: Python for solar physicists
SunPy is a data analysis toolkit which provides the necessary software for analyzing solar and heliospheric datasets in Python. SunPy …
Efficient stochastic Hessian estimation for full waveform inversion
In this abstract we present a method that allows arbitrary elements of the approximate Hessian to be estimated simultaneously. …
A phase field method for tomographic reconstruction from limited data
Classical tomographic reconstruction methods fail for problems in which there is extreme temporal and spatial sparsity in the measured …
Numerical methods for solar tomography in the STEREO era
In this thesis, we propose several advances in the numerical and computational algorithms that are used to determine tomographic …
A robust null space method for linear equality constrained state estimation
We present a robust null space method for linear equality constrained state space estimation. Exploiting a degeneracy in the estimator …
Dynamic three-dimensional tomography of the solar corona
Empirical, three-dimensional electron-density maps of the solar corona can be tomographically reconstructed using polarized-brightness …
Multiresolution analysis of active region magnetic structure and its correlation with the mount wilson classification and flaring activity
Two different multiresolution analyses are used to decompose the structure of active-region magnetic flux into concentrations of …
Multiscale Analysis of Active Region Evolution
Flows in the photosphere of solar active regions are turbulent in nature. Because magnetic fields are frozen into the plasma on the …
Multifractal Properties of Evolving Active Regions
Magnetohydrodynamic turbulence is thought to be responsible for producing complex, multiscale magnetic field distributions in solar …
Instant messaging (IM) literacy in the workplace
This chapter discusses instant messaging (IM) as a valuable digital tool that has influenced business communication practices at least …
Solar Activity Monitoring
SOLAR ACTIVITY MONITORING Here we review the current status of autonomous solar monitoring techniques of relevance to space weather. …
The Complex Sun: Turbulence and Complexity of the Solar atmosphere
It is already understood that simple monofractal and Fourier studies of magnetic field data give a good, though imperfect, …