SPLATT: Efficient and Parallel Sparse Tensor-Matrix Multiplication

Multi-dimensional arrays, or tensors, are increasingly found in fields such as signal processing and recommender systems. Real-world tensors can be enormous in size and often very sparse. There is a need for efficient, high-performance tools capable of processing the massive sparse tensors of today...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Proceedings - IEEE International Parallel and Distributed Processing Symposium s. 61 - 70
Hlavní autori: Smith, Shaden, Ravindran, Niranjay, Sidiropoulos, Nicholas D., Karypis, George
Médium: Konferenčný príspevok..
Jazyk:English
Vydavateľské údaje: IEEE 01.05.2015
Predmet:
Algorithm design and analysis
CANDECOMP
Context
CPD
Data structures
Memory management
PARAFAC
parallel
Parallel processing
Sparse matrices
Sparse tensors
Tensile stress
ISSN:1530-2075
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Multi-dimensional arrays, or tensors, are increasingly found in fields such as signal processing and recommender systems. Real-world tensors can be enormous in size and often very sparse. There is a need for efficient, high-performance tools capable of processing the massive sparse tensors of today and the future. This paper introduces SPLATT, a C library with shared-memory parallelism for three-mode tensors. SPLATT contains algorithmic improvements over competing state of the art tools for sparse tensor factorization. SPLATT has a fast, parallel method of multiplying a matricide tensor by a Khatri-Rao product, which is a key kernel in tensor factorization methods. SPLATT uses a novel data structure that exploits the sparsity patterns of tensors. This data structure has a small memory footprint similar to competing methods and allows for the computational improvements featured in our work. We also present a method of finding cache-friendly reordering and utilizing them with a novel form of cache tiling. To our knowledge, this is the first work to investigate reordering and cache tiling in this context. SPLATT averages almost 30x speedup compared to our baseline when using 16 threads and reaches over 80x speedup on NELL-2.
ISSN:1530-2075
DOI:10.1109/IPDPS.2015.27