An official website of the United States government
We implement two novel algorithms for sparse-matrix dense-matrix multiplication (SpMM) on the GPU. Our algorithms expect the sparse input in the popular compressed-sparse-row (CSR) format and thus do not require expensive format conversion. While previous SpMM work concentrates on thread-level parallelism, we additionally focus on latency hiding with instruction-level parallelism and load-balancing. We show, both theoretically and experimentally, that the proposed SpMM is a better fit for the GPU than previous approaches. We identify a key memory access pattern that allows efficient access into both input and output matrices that is crucial to getting excellent performance on SpMM. By combining these two ingredients---(i) merge-based load-balancing and (ii) row-major coalesced memory access---we demonstrate a 4.1x peak speedup and a 31.7% geomean speedup over state-of-the-art SpMM implementations on real-world datasets.
more »
« less
Learning from Optimizing Matrix-Matrix Multiplication
We describe a learning process that uses one of the simplest examples, matrix-matrix multiplication, to illustrate issues that underlie parallel high-performance computing. It is accessible at multiple levels: simple enough to use early in a curriculum yet rich enough to benefit a more advanced software developer. A carefully designed and scaffolded set of exercises leads the learner from a naive implementation towards one that extracts parallelism at multiple levels, ranging from instruction level parallelism to multithreaded parallelism via OpenMP to distributed memory parallelism using MPI. The importance of effectively leveraging the memory hierarchy within and across nodes is exposed, as do the GotoBLAS and SUMMA algorithms. These materials will become part of a Massive Open Online Course (MOOC) to be offered in the future.
more »
« less
- Award ID(s):
- 1714091
- PAR ID:
- 10073894
- Date Published:
- Journal Name:
- NSF/TCPP Workshop on Parallel and Distributed Computing Education (EduPar-18)
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Finite-state automata serve as compute kernels for many application domains such as pattern matching and data analytics. Existing approaches on GPUs exploit three levels of parallelism in automata processing tasks: 1)~input stream level, 2)~automaton-level and 3)~state-level. Among these, only state-level parallelism is intrinsic to automata while the other two levels of parallelism depend on the number of automata and input streams to be processed. As GPU resources increase, a parallelism-limited automata processing task can underutilize GPU compute resources. To this end, we propose AsyncAP, a low-overhead approach that optimizes for both scalability and throughput. Our insight is that most automata processing tasks have an additional source of parallelism originating from the input symbols which has not been leveraged before. Making the matching process associated with the automata tasks asynchronous, i.e., parallel GPU threads start processing an input stream from different input locations instead of processing it serially, improves throughput significantly and scales with input length. When the task does not have enough parallelism to utilize all the GPU cores, detailed evaluation across 12 evaluated applications shows that AsyncAP achieves up to 58× speedup on average over the state-of-the-art GPU automata processing engine. When the tasks have enough parallelism to utilize GPU cores, AsyncAP still achieves 2.4× speedup.more » « less
-
We develop a distributed-memory parallel algorithm for performing batch updates on streaming graphs, where vertices and edges are continuously added or removed. Our algorithm leverages distributed sparse matrices as the core data structures, utilizing equivalent sparse matrix operations to execute graph updates. By reducing unnecessary communication among processes and employing shared-memory parallelism, we accelerate updates of distributed graphs. Additionally, we maintain a balanced load in the output matrix by permuting the resultant matrix during the update process. We demonstrate that our streaming update algorithm is at least 25 times faster than alternative linear-algebraic methods and scales linearly up to 4,096 cores (32 nodes) on a Cray EX supercomputer.more » « less
-
We describe a randomized algorithm for producing a near-optimal hierarchical off-diagonal low-rank (HODLR) approximation to an n × n matrix A, accessible only though matrix-vector products with A and AT. We prove that, for the rank-k HODLR approximation problem, our method achieves a (1 + β )log(n )-optimal approximation in expected Frobenius norm using O (k log(n )/β3) matrix-vector products. In particular, the algorithm obtains a (1 + ∈ )-optimal approximation with O (k log4(n )/∈3) matrix-vector products, and for any constant c, an nc-optimal approximation with O (k log(n )) matrix-vector products. Apart from matrix-vector products, the additional computational cost of our method is just O (n poly(log(n ), k, β )). We complement the upper bound with a lower bound, which shows that any matrix-vector query algorithm requires at least Ω(k log(n ) + k/ε ) queries to obtain a (1 + ε )-optimal approximation. Our algorithm can be viewed as a robust version of widely used “peeling” methods for recovering HODLR matrices and is, to the best of our knowledge, the first matrix-vector query algorithm to enjoy theoretical worst- case guarantees for approximation by any hierarchical matrix class. To control the propagation of error between levels of hierarchical approximation, we introduce a new perturbation bound for low-rank approximation, which shows that the widely used Generalized Nyström method enjoys inherent stability when implemented with noisy matrix-vector products. We also introduce a novel randomly perforated matrix sketching method to further control the error in the peeling algorithm.more » « less
-
We consider the time and space required for quantum computers to solve a wide variety of problems involving matrices, many of which have only been analyzed classically in prior work. Our main results show that for a range of linear algebra problems—including matrix-vector product, matrix inversion, matrix multiplication and powering—existing classical time-space tradeoffs, several of which are tight for every space bound, also apply to quantum algorithms with at most a constant factor loss. For example, for almost all fixed matrices A, including the discrete Fourier transform (DFT) matrix, we prove that quantum circuits with at most T input queries and S qubits of memory require T = Ω(n^2/S) to compute matrix-vector product Ax for x ∈ {0, 1}^n. We similarly prove that matrix multiplication for n × n binary matrices requires T = Ω(n^3/√S). Because many of our lower bounds are matched by deterministic algorithms with the same time and space complexity, our results show that quantum computers cannot provide any asymptotic advantage for these problems with any space bound. We obtain matching lower bounds for the stronger notion of quantum cumulative memory complexity—the sum of the space per layer of a circuit. We also consider Boolean (i.e. AND-OR) matrix multiplication and matrix-vector products, improving the previous quantum time-space tradeoff lower bounds for n × n Boolean matrix multiplication to T = Ω(n^{2.5}/S^{1/4}) from T = Ω(n^{2.5}/S^{1/2}). Our improved lower bound for Boolean matrix multiplication is based on a new coloring argument that extracts more from the strong direct product theorem that was the basis for prior work. To obtain our tight lower bounds for linear algebra problems, we require much stronger bounds than strong direct product theorems. We obtain these bounds by adding a new bucketing method to the quantum recording-query technique of Zhandry that lets us apply classical arguments to upper bound the success probability of quantum circuits.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
