A note on the vectorization of scalar recursions (Q1819541)
From MaRDI portal
 | This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use this page instead for the normal view: A note on the vectorization of scalar recursions |
scientific article
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | A note on the vectorization of scalar recursions |
scientific article |
Statements
A note on the vectorization of scalar recursions (English)
0 references
1986
0 references
The authors present a comparison of scalar execution, product expansion (Neumann series) and cyclic reduction for the general first order recursion and Horner's scheme for 1 to 32 processors with and without chaining (it is not clear if bank conflicts for the cyclic reduction have been included). The optimal range for each method becomes visible.
0 references
parallel pipeline processors
0 references
recursive doubling
0 references
algorithmic complexity
0 references
vectorization
0 references
scalar execution
0 references
product expansion
0 references
Neumann series
0 references
cyclic reduction
0 references
first order recursion
0 references
Horner's scheme
0 references
bank conflicts
0 references
optimal range
0 references
