Search

We propose a novel approach to iterated sparse matrix dense matrix multiplication, a fundamental computational kernel in scientific computing and graph neural network training. In cases where matrix sizes exceed the memory of a single compute node, data transfer becomes a bottleneck. An approach based on dense matrix multiplication algorithms leads to suboptimal scalability and fails to exploit the sparsity in the problem. To address these challenges, we propose decomposing the sparse matrix into a small number of highly structured matrices called arrow matrices, which are connected by permutations. Our approach enables communication-avoiding multiplications, achieving a polynomial reduction in communication volume per iteration for matrices corresponding to planar graphs and other minor-excluded families of graphs. Our evaluation demonstrates that our approach outperforms a state-of-the-art method for sparse matrix multiplication on matrices with hundreds of millions of rows, offering near-linear strong and weak scaling.

The current hardware landscape and application scale is driving performance engineers towards writing bespoke optimizations. Verifying such optimizations, and generating minimal failing cases, is important for robustness in the face of changing program conditions, such as inputs and sizes. However, isolation of minimal test-cases from existing applications and generating new configurations are often difficult due to side effects on the system state, mostly related to dataflow. This paper introduces FuzzyFlow: a fault localization and test case extraction framework designed to test program optimizations. We leverage dataflow program representations to capture a fully reproducible system state and area-of-effect for optimizations to enable fast checking for semantic equivalence. To reduce testing time, we design an algorithm for minimizing test inputs, trading off memory for recomputation. We demonstrate FuzzyFlow on example use cases in real-world applications where the approach provides up to 528 times faster optimization testing and debugging compared to traditional approaches.

Multilinear algebra kernel performance on modern massively-parallel systems is determined mainly by data movement. However, deriving data movement-optimal distributed schedules for programs with many high-dimensional inputs is a notoriously hard problem. State-of-the-art libraries rely on heuristics and often fall back to suboptimal tensor folding and BLAS calls. We present Deinsum, an automated framework for distributed multilinear algebra computations expressed in Einstein notation, based on rigorous mathematical tools to address this problem. Our framework automatically derives data movement-optimal tiling and generates corresponding distributed schedules, further optimizing the performance of local computations by increasing their arithmetic intensity. To show the benefits of our approach, we test it on two important tensor kernel classes: Matricized Tensor Times Khatri-Rao Products and Tensor Times Matrix chains. We show performance results and scaling on the Piz Daint supercomputer, with up to 19x speedup over state-of-the-art solutions on 512 nodes.

Numerical codes that require arbitrary precision floating point (APFP) numbers for their core computation are dominated by elementary arithmetic operations due to the super-linear complexity of multiplication in the number of mantissa bits. APFP computations on conventional software-based architectures are made exceedingly expensive by the lack of native hardware support, requiring elementary operations to be emulated using instructions operating on machine-word-sized blocks. In this work, we show how APFP multiplication on compile-time fixed-precision operands can be implemented as deep FPGA pipelines with a recursively defined Karatsuba decomposition on top of native DSP multiplication. When comparing our design implemented on an Alveo U250 accelerator to a dual-socket 36-core Xeon node running the GNU Multiple Precision Floating-Point Reliable (MPFR) library, we achieve a 9.8x speedup at 4.8 GOp/s for 512-bit multiplication, and a 5.3x speedup at 1.2 GOp/s for 1024-bit multiplication, corresponding to the throughput of more than 351x and 191x CPU cores, respectively. We apply this architecture to general matrix-matrix multiplication, yielding a 10x speedup at 2.0 GOp/s over the Xeon node, equivalent to more than 375x CPU cores, effectively allowing a single FPGA to replace a small CPU cluster. Due to the significant dependence of some numerical codes on APFP, such as semidefinite program solvers, we expect these gains to translate into real-world speedups. Our configurable and flexible HLS-based code provides as high-level software interface for plug-and-play acceleration, published as an open source project.

Matrix factorizations are among the most important building blocks of scientific computing. State-of-the-art libraries, however, are not communication-optimal, underutilizing current parallel architectures. We present novel algorithms for Cholesky and LU factorizations that utilize an asymptotically communication-optimal 2.5D decomposition. We first establish a theoretical framework for deriving parallel I/O lower bounds for linear algebra kernels, and then utilize its insights to derive Cholesky and LU schedules, both communicating N^3/(P*sqrt(M)) elements per processor, where M is the local memory size. The empirical results match our theoretical analysis: our implementations communicate significantly less than Intel MKL, SLATE, and the asymptotically communication-optimal CANDMC and CAPITAL libraries. Our code outperforms these state-of-the-art libraries in almost all tested scenarios, with matrix sizes ranging from 2,048 to 262,144 on up to 512 CPU nodes of the Piz Daint supercomputer, decreasing the time-to-solution by up to three times. Our code is ScaLAPACK-compatible and available as an open-source library., Comment: 15 pages (including references), 11 figures. arXiv admin note: substantial text overlap with arXiv:2010.05975

Python has become the de facto language for scientific computing. Programming in Python is highly productive, mainly due to its rich science-oriented software ecosystem built around the NumPy module. As a result, the demand for Python support in High Performance Computing (HPC) has skyrocketed. However, the Python language itself does not necessarily offer high performance. In this work, we present a workflow that retains Python's high productivity while achieving portable performance across different architectures. The workflow's key features are HPC-oriented language extensions and a set of automatic optimizations powered by a data-centric intermediate representation. We show performance results and scaling across CPU, GPU, FPGA, and the Piz Daint supercomputer (up to 23,328 cores), with 2.47x and 3.75x speedups over previous-best solutions, first-ever Xilinx and Intel FPGA results of annotated Python, and up to 93.16% scaling efficiency on 512 nodes.

Determining I/O lower bounds is a crucial step in obtaining communication-efficient parallel algorithms, both across the memory hierarchy and between processors. Current approaches either study specific algorithms individually, disallow programmatic motifs such as recomputation, or produce asymptotic bounds that exclude important constants. We propose a novel approach for obtaining precise I/O lower bounds on a general class of programs, which we call Simple Overlap Access Programs (SOAP). SOAP analysis covers a wide variety of algorithms, from ubiquitous computational kernels to full scientific computing applications. Using the red-blue pebble game and combinatorial methods, we are able to bound the I/O of the SOAP-induced Computational Directed Acyclic Graph (CDAG), taking into account multiple statements, input/output reuse, and optimal tiling. To deal with programs that are outside of our representation (e.g., non-injective access functions), we describe methods to approximate them with SOAP. To demonstrate our method, we analyze 38 different applications, including kernels from the Polybench benchmark suite, deep learning operators, and -- for the first time -- applications in unstructured physics simulations, numerical weather prediction stencil compositions, and full deep neural networks. We derive tight I/O bounds for several linear algebra kernels, such as Cholesky decomposition, improving the existing reported bounds by a factor of two. For stencil applications, we improve the existing bounds by a factor of up to 14. We implement our method as an open-source tool, which can derive lower bounds directly from provided C code., Comment: 13 pages, 4 figures, published at Proceedings of the 33rd ACM Symposium on Parallelism in Algorithms and Architectures (SPAA'21)

Dense linear algebra kernels, such as linear solvers or tensor contractions, are fundamental components of many scientific computing applications. In this work, we present a novel method of deriving parallel I/O lower bounds for this broad family of programs. Based on the X-partitioning abstraction, our method explicitly captures inter-statement dependencies. Applying our analysis to LU factorization, we derive COnfLUX, an LU algorithm with the parallel I/O cost of $N^3 / (P \sqrt{M})$ communicated elements per processor -- only $1/3\times$ over our established lower bound. We evaluate COnfLUX on various problem sizes, demonstrating empirical results that match our theoretical analysis, communicating asymptotically less than Cray ScaLAPACK or SLATE, and outperforming the asymptotically-optimal CANDMC library. Running on $1$,$024$ nodes of Piz Daint, COnfLUX communicates 1.6$\times$ less than the second-best implementation and is expected to communicate 2.1$\times$ less on a full-scale run on Summit., Comment: 13 pages without references, 12 figures, submitted to PPoPP 2021: 26th ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming

The computational efficiency of a state of the art ab initio quantum transport (QT) solver, capable of revealing the coupled electro-thermal properties of atomically-resolved nano-transistors, has been improved by up to two orders of magnitude through a data centric reorganization of the application. The approach yields coarse-and fine-grained data-movement characteristics that can be used for performance and communication modeling, communication-avoidance, and dataflow transformations. The resulting code has been tuned for two top-6 hybrid supercomputers, reaching a sustained performance of 85.45 Pflop/s on 4,560 nodes of Summit (42.55% of the peak) in double precision, and 90.89 Pflop/s in mixed precision. These computational achievements enable the restructured QT simulator to treat realistic nanoelectronic devices made of more than 10,000 atoms within a 14$\times$ shorter duration than the original code needs to handle a system with 1,000 atoms, on the same number of CPUs/GPUs and with the same physical accuracy., Comment: 13 pages, 13 figures, SC19

Designing efficient cooling systems for integrated circuits (ICs) relies on a deep understanding of the electro-thermal properties of transistors. To shed light on this issue in currently fabricated FinFETs, a quantum mechanical solver capable of revealing atomically-resolved electron and phonon transport phenomena from first-principles is required. In this paper, we consider a global, data-centric view of a state-of-the-art quantum transport simulator to optimize its execution on supercomputers. The approach yields coarse- and fine-grained data-movement characteristics, which are used for performance and communication modeling, communication-avoidance, and data-layout transformations. The transformations are tuned for the Piz Daint and Summit supercomputers, where each platform requires different caching and fusion strategies to perform optimally. The presented results make ab initio device simulation enter a new era, where nanostructures composed of over 10,000 atoms can be investigated at an unprecedented level of accuracy, paving the way for better heat management in next-generation ICs., Comment: 12 pages, 18 figures, SC19

The ubiquity of accelerators in high-performance computing has driven programming complexity beyond the skill-set of the average domain scientist. To maintain performance portability in the future, it is imperative to decouple architecture-specific programming paradigms from the underlying scientific computations. We present the Stateful DataFlow multiGraph (SDFG), a data-centric intermediate representation that enables separating program definition from its optimization. By combining fine-grained data dependencies with high-level control-flow, SDFGs are both expressive and amenable to program transformations, such as tiling and double-buffering. These transformations are applied to the SDFG in an interactive process, using extensible pattern matching, graph rewriting, and a graphical user interface. We demonstrate SDFGs on CPUs, GPUs, and FPGAs over various motifs --- from fundamental computational kernels to graph analytics. We show that SDFGs deliver competitive performance, allowing domain scientists to develop applications naturally and port them to approach peak hardware performance without modifying the original scientific code.

We introduce Deep500: the first customizable benchmarking infrastructure that enables fair comparison of the plethora of deep learning frameworks, algorithms, libraries, and techniques. The key idea behind Deep500 is its modular design, where deep learning is factorized into four distinct levels: operators, network processing, training, and distributed training. Our evaluation illustrates that Deep500 is customizable (enables combining and benchmarking different deep learning codes) and fair (uses carefully selected metrics). Moreover, Deep500 is fast (incurs negligible overheads), verifiable (offers infrastructure to analyze correctness), and reproducible. Finally, as the first distributed and reproducible benchmarking system for deep learning, Deep500 provides software infrastructure to utilize the most powerful supercomputers for extreme-scale workloads., Comment: Accepted to IPDPS 2019