On BLAS Operations with Recursively Stored Sparse Matrices

Recently, we have proposed a recursive partitioning based layout for multi-core computations on sparse matrices. Based on positive results of our initial experiments with matrix-vector multiplication, we discuss how this storage format can be utilized across a range of BLAS-style matrix operations.

Uloženo v:
Podrobná bibliografie
Vydáno v:2010 12th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing s. 49 - 56
Hlavní autoři: Martone, M, Filippone, S, Paprzycki, M, Tucci, S
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.09.2010
Témata:
Artificial neural networks
blas
Instruction sets
Kernel
Matrix decomposition
multicore
parse matrix
Partitioning algorithms
recursive
recursive CSR
Sparse matrices
spmv
Symmetric matrices
ISBN:1424498163, 9781424498161
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Recently, we have proposed a recursive partitioning based layout for multi-core computations on sparse matrices. Based on positive results of our initial experiments with matrix-vector multiplication, we discuss how this storage format can be utilized across a range of BLAS-style matrix operations.
ISBN:1424498163
9781424498161
DOI:10.1109/SYNASC.2010.72