arXiv subjects

Edgar Solomonik

Publications and source records attributed to Edgar Solomonik.

At least 37 records · Page 2Linked to original sources

arXiv2021

Parallel Minimum Spanning Forest Computation using Sparse Matrix Kernels

Formulations of graph algorithms using sparse linear algebra have yielded highly scalable distributed algorithms for problems such as connectivity and shortest path computation. We develop the first formulation of the Awerbuch-Shiloach parallel minimum spanning forest (MSF) algorithm using linear algebra primitives. We introduce a multilinear kernel that operates on an adjacency matrix and two vectors. This kernel updates graph vertices by simultaneously using information from both adjacent edges and vertices. In addition, we explore optimizations to accelerate the shortcutting step in the Awerbuch-Shiloach algorithm. We implement this MSF algorithm with Cyclops, a distributed-memory library for generalized sparse tensor algebra. We analyze the parallel scalability of our implementation on the Stampede2 supercomputer.

cs.DC↗

arXiv2021

Fast and Accurate Randomized Algorithms for Low-rank Tensor Decompositions

Low-rank Tucker and CP tensor decompositions are powerful tools in data analytics. The widely used alternating least squares (ALS) method, which solves a sequence of over-determined least squares subproblems, is costly for large and sparse tensors. We propose a fast and accurate sketched ALS algorithm for Tucker decomposition, which solves a sequence of sketched rank-constrained linear least squares subproblems. Theoretical sketch size upper bounds are provided to achieve $O(ε)$ relative error for each subproblem with two sketching techniques, TensorSketch and leverage score sampling. Experimental results show that this new ALS algorithm, combined with a new initialization scheme based on randomized range finder, yields up to $22.0\%$ relative decomposition residual improvement compared to the state-of-the-art sketched randomized algorithm for Tucker decomposition of various synthetic and real datasets. This Tucker-ALS algorithm is further used to accelerate CP decomposition, by using randomized Tucker compression followed by CP decomposition of the Tucker core tensor. Experimental results show that this algorithm not only converges faster, but also yields more accurate CP decompositions.

math.NA↗

arXiv2021

Communication Lower Bounds of Bilinear Algorithms for Symmetric Tensor Contractions

We introduce a new theoretical framework for deriving lower bounds on data movement in bilinear algorithms. Bilinear algorithms are a general representation of fast algorithms for bilinear functions, which include computation of matrix multiplication, convolution, and symmetric tensor contractions. A bilinear algorithm is described by three matrices. Our communication lower bounds are based on quantifying the minimal matrix ranks of matching subsets of columns of these matrices. This infrastructure yields new communication lower bounds for symmetric tensor contraction algorithms, which provide qualitative new insights. Tensor symmetry (invariance under permutation of modes) is common to many applications of tensor computations (e.g., tensor representation of hypergraphs, analysis of high order moments in data, as well as tensors modelling interactions of electrons in computational chemistry). Tensor symmetry enables reduction in representation size as well as arithmetic cost of contractions by factors that scale with the number of equivalent permutations. However, we derive lower bounds showing that these arithmetic cost and memory reductions can necessitate increases in data movement by factors that scale with the size of the tensors.

cs.DC↗

arXiv2021

MTC: Multiresolution Tensor Completion from Partial and Coarse Observations

Existing tensor completion formulation mostly relies on partial observations from a single tensor. However, tensors extracted from real-world data are often more complex due to: (i) Partial observation: Only a small subset (e.g., 5%) of tensor elements are available. (ii) Coarse observation: Some tensor modes only present coarse and aggregated patterns (e.g., monthly summary instead of daily reports). In this paper, we are given a subset of the tensor and some aggregated/coarse observations (along one or more modes) and seek to recover the original fine-granular tensor with low-rank factorization. We formulate a coupled tensor completion problem and propose an efficient Multi-resolution Tensor Completion model (MTC) to solve the problem. Our MTC model explores tensor mode properties and leverages the hierarchy of resolutions to recursively initialize an optimization setup, and optimizes on the coupled system using alternating least squares. MTC ensures low computational and space complexity. We evaluate our model on two COVID-19 related spatio-temporal tensors. The experiments show that MTC could provide 65.20% and 75.79% percentage of fitness (PoF) in tensor completion with only 5% fine granular observations, which is 27.96% relative improvement over the best baseline. To evaluate the learned low-rank factors, we also design a tensor prediction task for daily and cumulative disease case predictions, where MTC achieves 50% in PoF and 30% relative improvements over the best baseline.

math.NA↗

arXiv2021

Distributed-Memory Tensor Completion for Generalized Loss Functions in Python using New Sparse Tensor Kernels

Tensor computations are increasingly prevalent numerical techniques in data science, but pose unique challenges for high-performance implementation. We provide novel algorithms and systems infrastructure which enable efficient parallel implementation of algorithms for tensor completion with generalized loss functions. Specifically, we consider alternating minimization, coordinate minimization, and a quasi-Newton (generalized Gauss-Newton) method. By extending the Cyclops library, we implement all of these methods in high-level Python syntax. To make possible tensor completion for very sparse tensors, we introduce new multi-tensor primitives, for which we provide specialized parallel implementations. We compare these routines to pairwise contraction of sparse tensors by reduction to hypersparse matrix formats, and find that the multi-tensor routines are more efficient in theoretical cost and execution time in experiments. We provide microbenchmarking results on the Stampede2 supercomputer to demonstrate the efficiency of the new primitives and Cyclops functionality. We then study the performance of the tensor completion methods for a synthetic tensor with 10 billion nonzeros and the Netflix dataset, considering both least squares and Poisson loss functions.

cs.DC↗

arXiv2021

Accelerating Alternating Least Squares for Tensor Decomposition by Pairwise Perturbation

The alternating least squares algorithm for CP and Tucker decomposition is dominated in cost by the tensor contractions necessary to set up the quadratic optimization subproblems. We introduce a novel family of algorithms that uses perturbative corrections to the subproblems rather than recomputing the tensor contractions. This approximation is accurate when the factor matrices are changing little across iterations, which occurs when alternating least squares approaches convergence. We provide a theoretical analysis to bound the approximation error. Our numerical experiments demonstrate that the proposed pairwise perturbation algorithms are easy to control and converge to minima that are as good as alternating least squares. The experimental results show improvements of up to 3.1X with respect to state-of-the-art alternating least squares approaches for various model tensor problems and real datasets.

math.NA↗

arXiv2021

Accelerating Distributed-Memory Autotuning via Statistical Analysis of Execution Paths

The prohibitive expense of automatic performance tuning at scale has largely limited the use of autotuning to libraries for shared-memory and GPU architectures. We introduce a framework for approximate autotuning that achieves a desired confidence in each algorithm configuration's performance by constructing confidence intervals to describe the performance of individual kernels (subroutines of benchmarked programs). Once a kernel's performance is deemed sufficiently predictable for a set of inputs, subsequent invocations are avoided and replaced with a predictive model of the execution time. We then leverage online execution path analysis to coordinate selective kernel execution and propagate each kernel's statistical profile. This strategy is effective in the presence of frequently-recurring computation and communication kernels, which is characteristic to algorithms in numerical linear algebra. We encapsulate this framework as part of a new profiling tool, Critter, that automates kernel execution decisions and propagates statistical profiles along critical paths of execution. We evaluate performance prediction accuracy obtained by our selective execution methods using state-of-the-art distributed-memory implementations of Cholesky and QR factorization on Stampede2, and demonstrate speed-ups of up to 7.1x with 98% prediction accuracy.

cs.DC↗

arXiv2020

AutoHOOT: Automatic High-Order Optimization for Tensors

High-order optimization methods, including Newton's method and its variants as well as alternating minimization methods, dominate the optimization algorithms for tensor decompositions and tensor networks. These tensor methods are used for data analysis and simulation of quantum systems. In this work, we introduce AutoHOOT, the first automatic differentiation (AD) framework targeting at high-order optimization for tensor computations. AutoHOOT takes input tensor computation expressions and generates optimized derivative expressions. In particular, AutoHOOT contains a new explicit Jacobian / Hessian expression generation kernel whose outputs maintain the input tensors' granularity and are easy to optimize. The expressions are then optimized by both the traditional compiler optimization techniques and specific tensor algebra transformations. Experimental results show that AutoHOOT achieves competitive CPU and GPU performance for both tensor decomposition and tensor network applications compared to existing AD software and other tensor computation libraries with manually written kernels. The tensor methods generated by AutoHOOT are also well-parallelizable, and we demonstrate good scalability on a distributed memory supercomputer.

cs.MS↗

arXiv2020

Communication-Efficient Jaccard Similarity for High-Performance Distributed Genome Comparisons

The Jaccard similarity index is an important measure of the overlap of two sets, widely used in machine learning, computational genomics, information retrieval, and many other areas. We design and implement SimilarityAtScale, the first communication-efficient distributed algorithm for computing the Jaccard similarity among pairs of large datasets. Our algorithm provides an efficient encoding of this problem into a multiplication of sparse matrices. Both the encoding and sparse matrix product are performed in a way that minimizes data movement in terms of communication and synchronization costs. We apply our algorithm to obtain similarity among all pairs of a set of large samples of genomes. This task is a key part of modern metagenomics analysis and an evergrowing need due to the increasing availability of high-throughput DNA sequencing data. The resulting scheme is the first to enable accurate Jaccard distance derivations for massive datasets, using largescale distributed-memory systems. We package our routines in a tool, called GenomeAtScale, that combines the proposed algorithm with tools for processing input sequences. Our evaluation on real data illustrates that one can use GenomeAtScale to effectively employ tens of thousands of processors to reach new frontiers in large-scale genomic and metagenomic analysis. While GenomeAtScale can be used to foster DNA research, the more general underlying SimilarityAtScale algorithm may be used for high-performance distributed similarity computations in other data analytics application domains.

cs.CE↗

arXiv2020

To Push or To Pull: On Reducing Communication and Synchronization in Graph Computations

We reduce the cost of communication and synchronization in graph processing by analyzing the fastest way to process graphs: pushing the updates to a shared state or pulling the updates to a private state.We investigate the applicability of this push-pull dichotomy to various algorithms and its impact on complexity, performance, and the amount of used locks, atomics, and reads/writes. We consider 11 graph algorithms, 3 programming models, 2 graph abstractions, and various families of graphs. The conducted analysis illustrates surprising differences between push and pull variants of different algorithms in performance, speed of convergence, and code complexity; the insights are backed up by performance data from hardware counters.We use these findings to illustrate which variant is faster for each algorithm and to develop generic strategies that enable even higher speedups. Our insights can be used to accelerate graph processing engines or libraries on both massively-parallel shared-memory machines as well as distributed-memory systems.

cs.DC↗

arXiv2020

Efficient parallel CP decomposition with pairwise perturbation and multi-sweep dimension tree

CP tensor decomposition with alternating least squares (ALS) is dominated in cost by the matricized-tensor times Khatri-Rao product (MTTKRP) kernel that is necessary to set up the quadratic optimization subproblems. State-of-art parallel ALS implementations use dimension trees to avoid redundant computations across MTTKRPs within each ALS sweep. In this paper, we propose two new parallel algorithms to accelerate CP-ALS. We introduce the multi-sweep dimension tree (MSDT) algorithm, which requires the contraction between an order N input tensor and the first-contracted input matrix once every (N-1)/N sweeps. This algorithm reduces the leading order computational cost by a factor of 2(N-1)/N relative to the best previously known approach. In addition, we introduce a more communication-efficient approach to parallelizing an approximate CP-ALS algorithm, pairwise perturbation. This technique uses perturbative corrections to the subproblems rather than recomputing the contractions, and asymptotically accelerates ALS. Our benchmark results show that the per-sweep time achieves 1.25X speed-up for MSDT and 1.94X speed-up for pairwise perturbation compared to the state-of-art dimension trees running on 1024 processors on the Stampede2 supercomputer.

cs.DC↗

arXiv2020

SlimSell: A Vectorizable Graph Representation for Breadth-First Search

Vectorization and GPUs will profoundly change graph processing. Traditional graph algorithms tuned for 32- or 64-bit based memory accesses will be inefficient on architectures with 512-bit wide (or larger) instruction units that are already present in the Intel Knights Landing (KNL) manycore CPU. Anticipating this shift, we propose SlimSell: a vectorizable graph representation to accelerate Breadth-First Search (BFS) based on sparse-matrix dense-vector (SpMV) products. SlimSell extends and combines the state-of-the-art SIMD-friendly Sell-C-sigma matrix storage format with tropical, real, boolean, and sel-max semiring operations. The resulting design reduces the necessary storage (by up to 50%) and thus pressure on the memory subsystem. We augment SlimSell with the SlimWork and SlimChunk schemes that reduce the amount of work and improve load balance, further accelerating BFS. We evaluate all the schemes on Intel Haswell multicore CPUs, the state-of-the-art Intel Xeon Phi KNL manycore CPUs, and NVIDIA Tesla GPUs. Our experiments indicate which semiring offers highest speedups for BFS and illustrate that SlimSell accelerates a tuned Graph500 BFS code by up to 33%. This work shows that vectorization can secure high-performance in BFS based on SpMV products; the proposed principles and designs can be extended to other graph algorithms.

cs.DC↗

arXiv2020

Efficient 2D Tensor Network Simulation of Quantum Systems

Simulation of quantum systems is challenging due to the exponential size of the state space. Tensor networks provide a systematically improvable approximation for quantum states. 2D tensor networks such as Projected Entangled Pair States (PEPS) are well-suited for key classes of physical systems and quantum circuits. However, direct contraction of PEPS networks has exponential cost, while approximate algorithms require computations with large tensors. We propose new scalable algorithms and software abstractions for PEPS-based methods, accelerating the bottleneck operation of contraction and refactorization of a tensor subnetwork. We employ randomized SVD with an implicit matrix to reduce cost and memory footprint asymptotically. Further, we develop a distributed-memory PEPS library and study accuracy and efficiency of alternative algorithms for PEPS contraction and evolution on the Stampede2 supercomputer. We also simulate a popular near-term quantum algorithm, the Variational Quantum Eigensolver (VQE), and benchmark Imaginary Time Evolution (ITE), which compute ground states of Hamiltonians.

cs.DC↗

arXiv2020

Pareto-Efficient Quantum Circuit Simulation Using Tensor Contraction Deferral

With the current rate of progress in quantum computing technologies, systems with more than 50 qubits will soon become reality. Computing ideal quantum state amplitudes for circuits of such and larger sizes is a fundamental step to assess both the correctness, performance, and scaling behavior of quantum algorithms and the fidelities of quantum devices. However, resource requirements for such calculations on classical computers grow exponentially. We show that deferring tensor contractions can extend the boundaries of what can be computed on classical systems. To demonstrate this technique, we present results obtained from a calculation of the complete set of output amplitudes of a universal random circuit with depth 27 in a 2D lattice of $7 \times 7$ qubits, and an arbitrarily selected slice of $2^{37}$ amplitudes of a universal random circuit with depth 23 in a 2D lattice of $8 \times 7$ qubits. Combining our methodology with other decomposition approaches found in the literature, we show that we can simulate $7 \times 7$-qubit random circuits to arbitrary depth by leveraging secondary storage. These calculations were thought to be impossible due to resource requirements.

quant-ph↗

arXiv2020

Distributed-Memory DMRG via Sparse and Dense Parallel Tensor Contractions

The Density Matrix Renormalization Group (DMRG) algorithm is a powerful tool for solving eigenvalue problems to model quantum systems. DMRG relies on tensor contractions and dense linear algebra to compute properties of condensed matter physics systems. However, its efficient parallel implementation is challenging due to limited concurrency, large memory footprint, and tensor sparsity. We mitigate these problems by implementing two new parallel approaches that handle block sparsity arising in DMRG, via Cyclops, a distributed memory tensor contraction library. We benchmark their performance on two physical systems using the Blue Waters and Stampede2 supercomputers. Our DMRG performance is improved by up to 5.9X in runtime and 99X in processing rate over ITensor, at roughly comparable computational resource use. This enables higher accuracy calculations via larger tensors for quantum state approximation. We demonstrate that despite having limited concurrency, DMRG is weakly scalable with the use of efficient parallel tensor contraction mechanisms.

cs.DC↗

arXiv2020

Derivation and Analysis of Fast Bilinear Algorithms for Convolution

The prevalence of convolution in applications within signal processing, deep neural networks, and numerical solvers has motivated the development of numerous fast convolution algorithms. In many of these problems, convolution is performed on terabytes or petabytes of data, so even constant factors of improvement can significantly reduce the computation time. We leverage the formalism of bilinear algorithms to describe and analyze all of the most popular approaches. This unified lens permits us to study the relationship between different variants of convolution as well as to derive error bounds and analyze the cost of the various algorithms. We provide new derivations, which predominantly leverage matrix and tensor algebra, to describe the Winograd family of convolution algorithms as well as reductions between 1D and multidimensional convolution. We provide cost and error bounds as well as experimental numerical studies. Our experiments for two of these algorithms, the overlap-add approach and Winograd convolution algorithm with polynomials of degree greater than one, show that fast convolution algorithms can rival the accuracy of the fast Fourier transform (FFT) without using complex arithmetic. These algorithms can be used for convolution problems with multidimensional inputs or for filters larger than size of four, extending the state-of-the-art in Winograd-based convolution algorithms.

math.NA↗

arXiv2020

Comparison of Accuracy and Scalability of Gauss-Newton and Alternating Least Squares for CP Decomposition

Alternating least squares is the most widely used algorithm for CP tensor decomposition. However, alternating least squares may exhibit slow or no convergence, especially when high accuracy is required. An alternative approach is to regard CP decomposition as a nonlinear least squares problem and employ Newton-like methods. Direct solution of linear systems involving an approximated Hessian is generally expensive. However, recent advancements have shown that use of an implicit representation of the linear system makes these methods competitive with alternating least squares. We provide the first parallel implementation of a Gauss-Newton method for CP decomposition, which iteratively solves linear least squares problems at each Gauss-Newton step. In particular, we leverage a formulation that employs tensor contractions for implicit matrix-vector products within the conjugate gradient method. The use of tensor contractions enables us to employ the Cyclops library for distributed-memory tensor computations to parallelize the Gauss-Newton approach with a high-level Python implementation. In addition, we propose a regularization scheme for Gauss-Newton method to improve convergence properties without any additional cost. We study the convergence of variants of the Gauss-Newton method relative to ALS for finding exact CP decompositions as well as approximate decompositions of real-world tensors. We evaluate the performance of sequential and parallel versions of both approaches, and study the parallel scalability on the Stampede2 supercomputer.

math.NA↗

arXiv2020

On Stability of Tensor Networks and Canonical Forms

Tensor networks such as matrix product states (MPS) and projected entangled pair states (PEPS) are commonly used to approximate quantum systems. These networks are optimized in methods such as DMRG or evolved by local operators. We provide bounds on the conditioning of tensor network representations to sitewise perturbations. These bounds characterize the extent to which local approximation error in the tensor sites of a tensor network can be amplified to error in the tensor it represents. In known tensor network methods, canonical forms of tensor network are used to minimize such error amplification. However, canonical forms are difficult to obtain for many tensor networks of interest. We quantify the extent to which error can be amplified in general tensor networks, yielding estimates of the benefit of the use of canonical forms. For the MPS and PEPS tensor networks, we provide simple forms on the worst-case error amplification. Beyond theoretical error bounds, we experimentally study the dependence of the error on the size of the network for perturbed random MPS tensor networks.

math.NA↗