Automatic data structure selection and transformation for sparse matrix computations

1996 ◽  
Vol 7 (2) ◽  
pp. 109-126 ◽  
Author(s):  
A.J.C. Bik ◽  
H.A.G. Wijshoff
Keyword(s):  
Data Structure ◽  
Sparse Matrix ◽  
Automatic Data ◽  
Matrix Computations ◽  
Structure Selection ◽  
Data Structure Selection

Automatic data structure selection in SETL

1979 ◽  
Author(s):  
Edmond Schonberg ◽  
Jacob T. Schwartz ◽  
Micha Sharir
Keyword(s):  
Data Structure ◽  
Automatic Data ◽  
Structure Selection ◽  
Data Structure Selection

Sparse triangular solves for ILU revisited: data layout crucial to better performance

2010 ◽  
Vol 25 (4) ◽  
pp. 386-391 ◽  
Author(s):  
Barry Smith ◽  
Hong Zhang
Keyword(s):  
Data Structure ◽  
Sparse Matrix ◽  
Memory System ◽  
Floating Point ◽  
Data Layout ◽  
Matrix Computations ◽  
Processor Utilization ◽  
Storage Format ◽  
Fast Access

A key to good processor utilization for sparse matrix computations is storing the data in the format that is most conducive to fast access by the memory system. In particular, for sparse matrix triangular solves the traditional compressed sparse matrix format is poor, and minor adjustments to the data structure can increase the processor utilization dramatically. Such adjustments involve storing the L and U factors separately and storing the U rows ‘backwards' so that they are accessed in a simple streaming fashion during the triangular solves. Changes to the PETSc libraries to use this modified storage format resulted in over twice the floating-point rate for some matrices. This improvement can be accounted for by a decrease in the cache misses and TLB (transaction lookaside buffer) misses in the modified code.

Sign in / Sign up

Export Citation Format

Share Document