arXiv subjects

Jan Hückelheim

Publications and source records attributed to Jan Hückelheim.

17 recordsLinked to original sources

arXiv2026

A Comprehensive Evaluation of Vertex Elimination Algorithms for Algorithmic Differentiation

The algorithmic differentiation (AD) of mathematical functions can be interpreted as a sequence of vertex eliminations in an underlying directed acyclic graph. The problem of determining a minimum-cost elimination ordering, which we call Optimal Vertex Elimination, is NP-complete. Consequently, much effort has been devoted to the design of heuristics. Many of these heuristics are widely believed to perform well in practice, but this hypothesis has so far been difficult to test due to the lack of scalable exact methods. We design and engineer new integer programming formulations for Optimal Vertex Eliminatioin and for a related objective we call Minimum Edge Count. Our implementations scale to graphs one-to-two orders of magnitude larger than existing techniques, enabling the assembly of a corpus of medium-sized graphs for which optimal solutions are known. This corpus facilitates a study of existing heuristics, confirming that on real data popular methods achieve high quality solutions. We also make several theoretical contributions. We give a tight analysis of the forward and reverse modes of AD, and extend our techniques to provide a simple algorithm for Optimal Vertex Elimination with approximation ratio parameterized by the size of a minimum source-sink separator. On the complexity side, we give the first approximation lower bounds for both problems.

cs.DS↗

arXiv2024

Profiling checkpointing schedules in adjoint ST-AD

Checkpointing is a cornerstone of data-flow reversal in adjoint algorithmic differentiation. Checkpointing is a storage/recomputation trade-off that can be applied at different levels, one of which being the call tree. We are looking for good placements of checkpoints onto the call tree of a given application, to reduce run time and memory footprint of its adjoint. There is no known optimal solution to this problem other than a combinatorial search on all placements. We propose a heuristics based on run-time profiling of the adjoint code. We describe implementation of this profiling tool in an existing source-transformation AD tool. We demonstrate the interest of this approach on test cases taken from the MITgcm ocean and atmospheric global circulation model. We discuss the limitations of our approach and propose directions to lift them.

cs.CL↗

arXiv2024

Challenges with Differentiable Quantum Dynamics

Differentiable quantum dynamics require automatic differentiation of a complex-valued initial value problem, which numerically integrates a system of ordinary differential equations from a specified initial condition, as well as the eigendecomposition of a matrix. We explored several automatic differentiation frameworks for these tasks, finding that no framework natively supports our application requirements. We therefore demonstrate a need for broader support of complex-valued, differentiable numerical integration in scientific computing libraries.

quant-ph↗

arXiv2024

Differentiating Through Linear Solvers

Computer programs containing calls to linear solvers are a known challenge for automatic differentiation. Previous publications advise against differentiating through the low-level solver implementation, and instead advocate for high-level approaches that express the derivative in terms of a modified linear system that can be solved with a separate solver call. Despite this ubiquitous advice, we are not aware of prior work comparing the accuracy of both approaches. With this article we thus empirically study a simple question: What happens if we ignore common wisdom, and differentiate through linear solvers?

cs.MS↗

arXiv2024

Parametric Sensitivities of a Wind-driven Baroclinic Ocean Using Neural Surrogates

Numerical models of the ocean and ice sheets are crucial for understanding and simulating the impact of greenhouse gases on the global climate. Oceanic processes affect phenomena such as hurricanes, extreme precipitation, and droughts. Ocean models rely on subgrid-scale parameterizations that require calibration and often significantly affect model skill. When model sensitivities to parameters can be computed by using approaches such as automatic differentiation, they can be used for such calibration toward reducing the misfit between model output and data. Because the SOMA model code is challenging to differentiate, we have created neural network-based surrogates for estimating the sensitivity of the ocean model to model parameters. We first generated perturbed parameter ensemble data for an idealized ocean model and trained three surrogate neural network models. The neural surrogates accurately predicted the one-step forward ocean dynamics, of which we then computed the parametric sensitivity.

physics.ao-ph↗

arXiv2023

Forward Gradients for Data-Driven CFD Wall Modeling

Computational Fluid Dynamics (CFD) is used in the design and optimization of gas turbines and many other industrial/ scientific applications. However, the practical use is often limited by the high computational cost, and the accurate resolution of near-wall flow is a significant contributor to this cost. Machine learning (ML) and other data-driven methods can complement existing wall models. Nevertheless, training these models is bottlenecked by the large computational effort and memory footprint demanded by back-propagation. Recent work has presented alternatives for computing gradients of neural networks where a separate forward and backward sweep is not needed and storage of intermediate results between sweeps is not required because an unbiased estimator for the gradient is computed in a single forward sweep. In this paper, we discuss the application of this approach for training a subgrid wall model that could potentially be used as a surrogate in wall-bounded flow CFD simulations to reduce the computational overhead while preserving predictive accuracy.

physics.flu-dyn↗

arXiv2023

Surrogate Neural Networks to Estimate Parametric Sensitivity of Ocean Models

Modeling is crucial to understanding the effect of greenhouse gases, warming, and ice sheet melting on the ocean. At the same time, ocean processes affect phenomena such as hurricanes and droughts. Parameters in the models that cannot be physically measured have a significant effect on the model output. For an idealized ocean model, we generated perturbed parameter ensemble data and trained surrogate neural network models. The neural surrogates accurately predicted the one-step forward dynamics, of which we then computed the parametric sensitivity.

physics.ao-ph↗

arXiv2023

Model Checking Race-freedom When "Sequential Consistency for Data-race-free Programs" is Guaranteed

Many parallel programming models guarantee that if all sequentially consistent (SC) executions of a program are free of data races, then all executions of the program will appear to be sequentially consistent. This greatly simplifies reasoning about the program, but leaves open the question of how to verify that all SC executions are race-free. In this paper, we show that with a few simple modifications, model checking can be an effective tool for verifying race-freedom. We explore this technique on a suite of C programs parallelized with OpenMP.

cs.PL↗

arXiv2023

Understanding Automatic Differentiation Pitfalls

Automatic differentiation, also known as backpropagation, AD, autodiff, or algorithmic differentiation, is a popular technique for computing derivatives of computer programs accurately and efficiently. Sometimes, however, the derivatives computed by AD could be interpreted as incorrect. These pitfalls occur systematically across tools and approaches. In this paper we broadly categorize problematic usages of AD and illustrate each category with examples such as chaos, time-averaged oscillations, discretizations, fixed-point loops, lookup tables, and linear solvers. We also review debugging techniques and their effectiveness in these situations. With this article we hope to help readers avoid unexpected behavior, detect problems more easily when they occur, and have more realistic expectations from AD tools.

math.NA↗

arXiv2022

Memory-Efficient Differentiable Programming for Quantum Optimal Control of Discrete Lattices

Quantum optimal control problems are typically solved by gradient-based algorithms such as GRAPE, which suffer from exponential growth in storage with increasing number of qubits and linear growth in memory requirements with increasing number of time steps. Employing QOC for discrete lattices reveals that these memory requirements are a barrier for simulating large models or long time spans. We employ a nonstandard differentiable programming approach that significantly reduces the memory requirements at the cost of a reasonable amount of recomputation. The approach exploits invertibility properties of the unitary matrices to reverse the computation during back-propagation. We utilize QOC software written in the differentiable programming framework JAX that implements this approach, and demonstrate its effectiveness for lattice gauge theory.

quant-ph↗

arXiv2022

TROPHY: Trust Region Optimization Using a Precision Hierarchy

We present an algorithm to perform trust-region-based optimization for nonlinear unconstrained problems. The method selectively uses function and gradient evaluations at different floating-point precisions to reduce the overall energy consumption, storage, and communication costs; these capabilities are increasingly important in the era of exascale computing. In particular, we are motivated by a desire to improve computational efficiency for massive climate models. We employ our method on two examples: the CUTEst test set and a large-scale data assimilation problem to recover wind fields from radar returns. Although this paper is primarily a proof of concept, we show that if implemented on appropriate hardware, the use of mixed-precision can significantly reduce the computational load compared with fixed-precision solvers.

math.OC↗

arXiv2021

Source-to-Source Automatic Differentiation of OpenMP Parallel Loops

This paper presents our work toward correct and efficient automatic differentiation of OpenMP parallel worksharing loops in forward and reverse mode. Automatic differentiation is a method to obtain gradients of numerical programs, which are crucial in optimization, uncertainty quantification, and machine learning. The computational cost to compute gradients is a common bottleneck in practice. For applications that are parallelized for multicore CPUs or GPUs using OpenMP, one also wishes to compute the gradients in parallel. We propose a framework to reason about the correctness of the generated derivative code, from which we justify our OpenMP extension to the differentiation model. We implement this model in the automatic differentiation tool Tapenade and present test cases that are differentiated following our extended differentiation procedure. Performance of the generated derivative programs in forward and reverse mode is better than sequential, although our reverse mode often scales worse than the input programs.

cs.MS↗

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

Automatic Differentiation for Adjoint Stencil Loops

Stencil loops are a common motif in computations including convolutional neural networks, structured-mesh solvers for partial differential equations, and image processing. Stencil loops are easy to parallelise, and their fast execution is aided by compilers, libraries, and domain-specific languages. Reverse-mode automatic differentiation, also known as algorithmic differentiation, autodiff, adjoint differentiation, or back-propagation, is sometimes used to obtain gradients of programs that contain stencil loops. Unfortunately, conventional automatic differentiation results in a memory access pattern that is not stencil-like and not easily parallelisable. In this paper we present a novel combination of automatic differentiation and loop transformations that preserves the structure and memory access pattern of stencil loops, while computing fully consistent derivatives. The generated loops can be parallelised and optimised for performance in the same way and using the same tools as the original computation. We have implemented this new technique in the Python tool PerforAD, which we release with this paper along with test cases derived from seismic imaging and computational fluid dynamics applications.

cs.DC↗

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↗

arXiv2018

High-level python abstractions for optimal checkpointing in inversion problems

Inversion and PDE-constrained optimization problems often rely on solving the adjoint problem to calculate the gradient of the objec- tive function. This requires storing large amounts of intermediate data, setting a limit to the largest problem that might be solved with a given amount of memory available. Checkpointing is an approach that can reduce the amount of memory required by redoing parts of the computation instead of storing intermediate results. The Revolve checkpointing algorithm o ers an optimal schedule that trades computational cost for smaller memory footprints. Integrat- ing Revolve into a modern python HPC code and combining it with code generation is not straightforward. We present an API that makes checkpointing accessible from a DSL-based code generation environment along with some initial performance gures with a focus on seismic applications.

cs.MS↗

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↗