Cyclops Tensor Framework: Reducing Communication and Eliminating Load Imbalance in Massively Parallel Contractions

Cyclops (cyclic-operations) Tensor Framework (CTF) 1 is a distributed library for tensor contractions. CTF aims to scale high-dimensional tensor contractions such as those required in the Coupled Cluster (CC) electronic structure method to massively-parallel supercomputers. The framework preserves t...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2013 IEEE 27th International Symposium on Parallel and Distributed Processing s. 813 - 824
Hlavní autoři: Solomonik, Edgar, Matthews, Devin, Hammond, Jeff R., Demmel, James
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.05.2013
Témata:
Chemistry
Clustering algorithms
communication-avoiding algorithms
Coupled Cluster
Cyclops
Equations
Indexes
Manganese
Program processors
Tensile stress
tensor contractions
ISBN:146736066X, 9781467360661
ISSN:1530-2075
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Cyclops (cyclic-operations) Tensor Framework (CTF) 1 is a distributed library for tensor contractions. CTF aims to scale high-dimensional tensor contractions such as those required in the Coupled Cluster (CC) electronic structure method to massively-parallel supercomputers. The framework preserves tensor structure by subdividing tensors cyclically, producing a regular parallel decomposition. An internal virtualization layer provides completely general mapping support while maintaining ideal load balance. The mapping framework decides on the best mapping for each tensor contraction at run-time via explicit calculations of memory usage and communication volume. CTF employs a general redistribution kernel, which transposes tensors of any dimension between arbitrary distributed layouts, yet touches each piece of data only once. Sequential symmetric contractions are reduced to matrix multiplication calls via tensor index transpositions and partial unpacking. The user-level interface elegantly expresses arbitrary-dimensional generalized tensor contractions in the form of a domain specific language. We demonstrate performance of CC with single and double excitations on 8192 nodes of Blue Gene/Q and show that CTF outperforms NWChem on Cray XE6 supercomputers for benchmarked systems.
ISBN:146736066X
9781467360661
ISSN:1530-2075
DOI:10.1109/IPDPS.2013.112