On Optimizing Operator Fusion Plans for Large-Scale Machine Learning in SystemML
January 02, 2018 Β· Declared Dead Β· π Proceedings of the VLDB Endowment
"No code URL or promise found in abstract"
Evidence collected by the PWNC Scanner
Authors
Matthias Boehm, Berthold Reinwald, Dylan Hutchison, Alexandre V. Evfimievski, Prithviraj Sen
arXiv ID
1801.00829
Category
cs.DB: Databases
Citations
61
Venue
Proceedings of the VLDB Endowment
Last Checked
3 months ago
Abstract
Many large-scale machine learning (ML) systems allow specifying custom ML algorithms by means of linear algebra programs, and then automatically generate efficient execution plans. In this context, optimization opportunities for fused operators---in terms of fused chains of basic operators---are ubiquitous. These opportunities include (1) fewer materialized intermediates, (2) fewer scans of input data, and (3) the exploitation of sparsity across chains of operators. Automatic operator fusion eliminates the need for hand-written fused operators and significantly improves performance for complex or previously unseen chains of operations. However, existing fusion heuristics struggle to find good fusion plans for complex DAGs or hybrid plans of local and distributed operations. In this paper, we introduce an optimization framework for systematically reason about fusion plans that considers materialization points in DAGs, sparsity exploitation, different fusion template types, as well as local and distributed operations. In detail, we contribute algorithms for (1) candidate exploration of valid fusion plans, (2) cost-based candidate selection, and (3) code generation of local and distributed operations over dense, sparse, and compressed data. Our experiments in SystemML show end-to-end performance improvements with optimized fusion plans of up to 21x compared to hand-written fused operators, with negligible optimization and code generation overhead.
Community Contributions
Found the code? Know the venue? Think something is wrong? Let us know!
π Similar Papers
In the same crypt β Databases
R.I.P.
π»
Ghosted
R.I.P.
π»
Ghosted
Untangling Blockchain: A Data Processing View of Blockchain Systems
R.I.P.
π»
Ghosted
Converting Static Image Datasets to Spiking Neuromorphic Datasets Using Saccades
R.I.P.
π»
Ghosted
BLOCKBENCH: A Framework for Analyzing Private Blockchains
R.I.P.
π»
Ghosted
Data Synthesis based on Generative Adversarial Networks
R.I.P.
π»
Ghosted
HoloClean: Holistic Data Repairs with Probabilistic Inference
Died the same way β π» Ghosted
R.I.P.
π»
Ghosted
Federated Learning: Strategies for Improving Communication Efficiency
R.I.P.
π»
Ghosted
In-Datacenter Performance Analysis of a Tensor Processing Unit
R.I.P.
π»
Ghosted
Deep Convolutional Neural Networks for Computer-Aided Detection: CNN Architectures, Dataset Characteristics and Transfer Learning
R.I.P.
π»
Ghosted