arXiv subjects

Gerard J. Gorman

Publications and source records attributed to Gerard J. Gorman.

At least 19 recordsLinked to original sources

arXiv2023

Learned multiphysics inversion with differentiable programming and machine learning

We present the Seismic Laboratory for Imaging and Modeling/Monitoring (SLIM) open-source software framework for computational geophysics and, more generally, inverse problems involving the wave-equation (e.g., seismic and medical ultrasound), regularization with learned priors, and learned neural surrogates for multiphase flow simulations. By integrating multiple layers of abstraction, our software is designed to be both readable and scalable. This allows researchers to easily formulate their problems in an abstract fashion while exploiting the latest developments in high-performance computing. We illustrate and demonstrate our design principles and their benefits by means of building a scalable prototype for permeability inversion from time-lapse crosswell seismic data, which aside from coupling of wave physics and multiphase flow, involves machine learning.

cs.MS↗

arXiv2021

PyBryt: auto-assessment and auto-grading for computational thinking

We continuously interact with computerized systems to achieve goals and perform tasks in our personal and professional lives. Therefore, the ability to program such systems is a skill needed by everyone. Consequently, computational thinking skills are essential for everyone, which creates a challenge for the educational system to teach these skills at scale and allow students to practice these skills. To address this challenge, we present a novel approach to providing formative feedback to students on programming assignments. Our approach uses dynamic evaluation to trace intermediate results generated by student's code and compares them to the reference implementation provided by their teachers. We have implemented this method as a Python library and demonstrate its use to give students relevant feedback on their work while allowing teachers to challenge their students' computational thinking skills.

cs.HC↗

arXiv2021

Lossy Checkpoint Compression in Full Waveform Inversion: a case study with ZFPv0.5.5 and the Overthrust Model

This paper proposes a new method that combines check-pointing methods with error-controlled lossy compression for large-scale high-performance Full-Waveform Inversion (FWI), an inverse problem commonly used in geophysical exploration. This combination can significantly reduce data movement, allowing a reduction in run time as well as peak memory. In the Exascale computing era, frequent data transfer (e.g., memory bandwidth, PCIe bandwidth for GPUs, or network) is the performance bottleneck rather than the peak FLOPS of the processing unit. Like many other adjoint-based optimization problems, FWI is costly in terms of the number of floating-point operations, large memory footprint during backpropagation, and data transfer overheads. Past work for adjoint methods has developed checkpointing methods that reduce the peak memory requirements during backpropagation at the cost of additional floating-point computations. Combining this traditional checkpointing with error-controlled lossy compression, we explore the three-way tradeoff between memory, precision, and time to solution. We investigate how approximation errors introduced by lossy compression of the forward solution impact the objective function gradient and final inverted solution. Empirical results from these numerical experiments indicate that high lossy-compression rates (compression factors ranging up to 100) have a relatively minor impact on convergence rates and the quality of the final solution.

physics.comp-ph↗

arXiv2020

Architecture and performance of Devito, a system for automated stencil computation

Stencil computations are a key part of many high-performance computing applications, such as image processing, convolutional neural networks, and finite-difference solvers for partial differential equations. Devito is a framework capable of generating highly-optimized code given symbolic equations expressed in Python, specialized in, but not limited to, affine (stencil) codes. The lowering process---from mathematical equations down to C++ code---is performed by the Devito compiler through a series of intermediate representations. Several performance optimizations are introduced, including advanced common sub-expressions elimination, tiling and parallelization. Some of these are obtained through well-established stencil optimizers, integrated in the back-end of the Devito compiler. The architecture of the Devito compiler, as well as the performance optimizations that are applied when generating code, are presented. The effectiveness of such performance optimizations is demonstrated using operators drawn from seismic imaging applications.

cs.MS↗

arXiv2019

Performance of Devito on HPC-Optimised ARM Processors

We evaluate the performance of Devito, a domain specific language (DSL) for finite differences on Arm ThunderX2 processors. Experiments with two common seismic computational kernels demonstrate that Arm processors can deliver competitive performance compared to other Intel Xeon processors.

cs.PF↗

arXiv2019

Devito (v3.1.0): an embedded domain-specific language for finite differences and geophysical exploration

We introduce Devito, a new domain-specific language for implementing high-performance finite difference partial differential equation solvers. The motivating application is exploration seismology where methods such as Full-Waveform Inversion and Reverse-Time Migration are used to invert terabytes of seismic data to create images of the earth's subsurface. Even using modern supercomputers, it can take weeks to process a single seismic survey and create a useful subsurface image. The computational cost is dominated by the numerical solution of wave equations and their corresponding adjoints. Therefore, a great deal of effort is invested in aggressively optimizing the performance of these wave-equation propagators for different computer architectures. Additionally, the actual set of partial differential equations being solved and their numerical discretization is under constant innovation as increasingly realistic representations of the physics are developed, further ratcheting up the cost of practical solvers. By embedding a domain-specific language within Python and making heavy use of SymPy, a symbolic mathematics library, we make it possible to develop finite difference simulators quickly using a syntax that strongly resembles the mathematics. The Devito compiler reads this code and applies a wide range of analysis to generate highly optimized and parallel code. This approach can reduce the development time of a verified and optimized solver from months to days.

cs.DM↗

arXiv2019

Automated Tiling of Unstructured Mesh Computations with Application to Seismological Modelling

Sparse tiling is a technique to fuse loops that access common data, thus increasing data locality. Unlike traditional loop fusion or blocking, the loops may have different iteration spaces and access shared datasets through indirect memory accesses, such as A[map[i]] -- hence the name "sparse". One notable example of such loops arises in discontinuous-Galerkin finite element methods, because of the computation of numerical integrals over different domains (e.g., cells, facets). The major challenge with sparse tiling is implementation -- not only is it cumbersome to understand and synthesize, but it is also onerous to maintain and generalize, as it requires a complete rewrite of the bulk of the numerical computation. In this article, we propose an approach to extend the applicability of sparse tiling based on raising the level of abstraction. Through a sequence of compiler passes, the mathematical specification of a problem is progressively lowered, and eventually sparse-tiled C for-loops are generated. Besides automation, we advance the state-of-the-art by introducing: a revisited, more efficient sparse tiling algorithm; support for distributed-memory parallelism; a range of fine-grained optimizations for increased run-time performance; implementation in a publicly-available library, SLOPE; and an in-depth study of the performance impact in Seigen, a real-world elastic wave equation solver for seismological problems, which shows speed-ups up to 1.28x on a platform consisting of 896 Intel Broadwell cores.

cs.CE↗

arXiv2018

Backpropagation for long sequences: beyond memory constraints with constant overheads

Naive backpropagation through time has a memory footprint that grows linearly in the sequence length, due to the need to store each state of the forward propagation. This is a problem for large networks. Strategies have been developed to trade memory for added computations, which results in a sublinear growth of memory footprint or computation overhead. In this work, we present a library that uses asynchronous storing and prefetching to move data to and from slow and cheap stor- age. The library only stores and prefetches states as frequently as possible without delaying the computation, and uses the optimal Revolve backpropagation strategy for the computations in between. The memory footprint of the backpropagation can thus be reduced to any size (e.g. to fit into DRAM), while the computational overhead is constant in the sequence length, and only depends on the ratio between compute and transfer times on a given hardware. We show in experiments that by exploiting asyncronous data transfer, our strategy is always at least as fast, and usually faster than the previously studied "optimal" strategies.

cs.DC↗

arXiv2017

Optimised finite difference computation from symbolic equations

Domain-specific high-productivity environments are playing an increasingly important role in scientific computing due to the levels of abstraction and automation they provide. In this paper we introduce Devito, an open-source domain-specific framework for solving partial differential equations from symbolic problem definitions by the finite difference method. We highlight the generation and automated execution of highly optimized stencil code from only a few lines of high-level symbolic Python for a set of scientific equations, before exploring the use of Devito operators in seismic inversion problems.

cs.MS↗

arXiv2016

Anisotropic mesh adaptation in Firedrake with PETSc DMPlex

Despite decades of research in this area, mesh adaptation capabilities are still rarely found in numerical simulation software. We postulate that the primary reason for this is lack of usability. Integrating mesh adaptation into existing software is difficult as non-trivial operators, such as error metrics and interpolation operators, are required, and integrating available adaptive remeshers is not straightforward. Our approach presented here is to first integrate Pragmatic, an anisotropic mesh adaptation library, into DMPlex, a PETSc object that manages unstructured meshes and their interactions with PETSc's solvers and I/O routines. As PETSc is already widely used, this will make anisotropic mesh adaptation available to a much larger community. As a demonstration of this we describe the integration of anisotropic mesh adaptation into Firedrake, an automated Finite Element based system for the portable solution of partial differential equations which already uses PETSc solvers and I/O via DMPlex. We present a proof of concept of this integration with a three-dimensional advection test case.

math.NA↗

arXiv2016

Experiences with efficient methodologies for teaching computer programming to geoscientists

Computer programming was once thought of as a skill required only by professional software developers. But today, given the ubiquitous nature of computation and data science it is quickly becoming necessary for all scientists and engineers to have at least a basic knowledge of how to program. Teaching how to program, particularly to those students with little or no computing background, is well-known to be a difficult task. However, there is also a wealth of evidence-based teaching practices for teaching programming skills which can be applied to greatly improve learning outcomes and the student experience. Adopting these practices naturally gives rise to greater learning efficiency - this is critical if programming is to be integrated into an already busy geoscience curriculum. This paper considers an undergraduate computer programming course, run during the last 5 years in the Department of Earth Science and Engineering at Imperial College London. The teaching methodologies that were used each year are discussed alongside the challenges that were encountered, and how the methodologies affected student performance. Anonymised student marks and feedback are used to highlight this, and also how the adjustments made to the course eventually resulted in a highly effective learning environment.

cs.CY↗

arXiv2015

Shoreline and Bathymetry Approximation in Mesh Generation for Tidal Renewable Simulations

Due to the fractal nature of the domain geometry in geophysical flow simulations, a completely accurate description of the domain in terms of a computational mesh is frequently deemed infeasible. Shoreline and bathymetry simplification methods are used to remove small scale details in the geometry, particularly in areas away from the region of interest. To that end, a novel method for shoreline and bathymetry simplification is presented. Existing shoreline simplification methods typically remove points if the resultant geometry satisfies particular geometric criteria. Bathymetry is usually simplified using traditional filtering techniques, that remove unwanted Fourier modes. Principal Component Analysis (PCA) has been used in other fields to isolate small-scale structures from larger scale coherent features in a robust way, underpinned by a rigorous but simple mathematical framework. Here we present a method based on principal component analysis aimed towards simplification of shorelines and bathymetry. We present the algorithm in detail and show simplified shorelines and bathymetry in the wider region around the North Sea. Finally, the methods are used in the context of unstructured mesh generation aimed at tidal resource assessment simulations in the coastal regions around the UK.

cs.CG↗

arXiv2015

Efficient mesh management in Firedrake using PETSc-DMPlex

The use of composable abstractions allows the application of new and established algorithms to a wide range of problems while automatically inheriting the benefits of well-known performance optimisations. This work highlights the composition of the PETSc DMPlex domain topology abstraction with the Firedrake automated finite element system to create a PDE solving environment that combines expressiveness, flexibility and high performance. We describe how Firedrake utilises DMPlex to provide the indirection maps required for finite element assembly, while supporting various mesh input formats and runtime domain decomposition. In particular, we describe how DMPlex and its accompanying data structures allow the generic creation of user-defined discretisations, while utilising data layout optimisations that improve cache coherency and ensure overlapped communication during assembly computation.

cs.MS↗

arXiv2015

Unstructured Overlapping Mesh Distribution in Parallel

We present a simple mathematical framework and API for parallel mesh and data distribution, load balancing, and overlap generation. It relies on viewing the mesh as a Hasse diagram, abstracting away information such as cell shape, dimension, and coordinates. The high level of abstraction makes our interface both concise and powerful, as the same algorithm applies to any representable mesh, such as hybrid meshes, meshes embedded in higher dimension, and overlapped meshes in parallel. We present evidence, both theoretical and experimental, that the algorithms are scalable and efficient. A working implementation can be found in the latest release of the PETSc libraries.

cs.MS↗

arXiv2015

An Interrupt-Driven Work-Sharing For-Loop Scheduler

In this paper we present a parallel for-loop scheduler which is based on work-stealing principles but runs under a completely cooperative scheme. POSIX signals are used by idle threads to interrupt left-behind workers, which in turn decide what portion of their workload can be given to the requester. We call this scheme Interrupt-Driven Work-Sharing (IDWS). This article describes how IDWS works, how it can be integrated into any POSIX-compliant OpenMP implementation and how a user can manually replace OpenMP parallel for-loops with IDWS in existing POSIX-compliant C++ applications. Additionally, we measure its performance using both a synthetic benchmark with varying distributions of workload across the iteration space and a real-life application on Sandy Bridge and Xeon Phi systems. Regardless the workload distribution and the underlying hardware, IDWS is always the best or among the best-performing strategies, providing a good all-around solution to the scheduling-choice dilemma.

cs.DC↗

arXiv2015

Flexible, Scalable Mesh and Data Management using PETSc DMPlex

Designing a scientific software stack to meet the needs of the next-generation of mesh-based simulation demands, not only scalable and efficient mesh and data management on a wide range of platforms, but also an abstraction layer that makes it useful for a wide range of application codes. Common utility tasks, such as file I/O, mesh distribution, and work partitioning, should be delegated to external libraries in order to promote code re-use, extensibility and software interoperability. In this paper we demonstrate the use of PETSc's DMPlex data management API to perform mesh input and domain partitioning in Fluidity, a large scale CFD application. We demonstrate that raising the level of abstraction adds new functionality to the application code, such as support for additional mesh file formats and mesh re- ordering, while improving simulation startup cost through more efficient mesh distribution. Moreover, the separation of concerns accomplished through this interface shifts critical performance and interoperability issues, such as scalable I/O and file format support, to a widely used and supported open source community library, improving the sustainability, performance, and functionality of Fluidity.

cs.MS↗

arXiv2015

Thread Parallelism for Highly Irregular Computation in Anisotropic Mesh Adaptation

Thread-level parallelism in irregular applications with mutable data dependencies presents challenges because the underlying data is extensively modified during execution of the algorithm and a high degree of parallelism must be realized while keeping the code race-free. In this article we describe a methodology for exploiting thread parallelism for a class of graph-mutating worklist algorithms, which guarantees safe parallel execution via processing in rounds of independent sets and using a deferred update strategy to commit changes in the underlying data structures. Scalability is assisted by atomic fetch-and-add operations to create worklists and work-stealing to balance the shared-memory workload. This work is motivated by mesh adaptation algorithms, for which we show a parallel efficiency of 60% and 50% on Intel(R) Xeon(R) Sandy Bridge and AMD Opteron(tm) Magny-Cours systems, respectively, using these techniques.

cs.DC↗

arXiv2014

PyRDM: A Python-based library for automating the management and online publication of scientific software and data

The recomputability and reproducibility of results from scientific software requires access to both the source code and all associated input and output data. However, the full collection of these resources often does not accompany the key findings published in journal articles, thereby making it difficult or impossible for the wider scientific community to verify the correctness of a result or to build further research on it. This paper presents a new Python-based library, PyRDM, whose functionality aims to automate the process of sharing the software and data via online, citable repositories such as Figshare. The library is integrated into the workflow of an open-source computational fluid dynamics package, Fluidity, to demonstrate an example of its usage.

cs.CE↗