Using non-canonical array layouts in dense matrix operations
JR Herrero, JJ Navarro - International Workshop on Applied Parallel …, 2006 - Springer
JR Herrero, JJ Navarro
International Workshop on Applied Parallel Computing, 2006•SpringerWe present two implementations of dense matrix multiplication based on two different non-
canonical array layouts: one based on a hypermatrix data structure (HM) where data
submatrices are stored using a recursive layout; the other based on a simple block data
layout with square blocks (SB) where blocks are arranged in column-major order. We show
that the iterative code using SB outperforms a recursive code using HM and obtains
competitive results on a variety of platforms.
canonical array layouts: one based on a hypermatrix data structure (HM) where data
submatrices are stored using a recursive layout; the other based on a simple block data
layout with square blocks (SB) where blocks are arranged in column-major order. We show
that the iterative code using SB outperforms a recursive code using HM and obtains
competitive results on a variety of platforms.
Abstract
We present two implementations of dense matrix multiplication based on two different non-canonical array layouts: one based on a hypermatrix data structure (HM) where data submatrices are stored using a recursive layout; the other based on a simple block data layout with square blocks (SB) where blocks are arranged in column-major order. We show that the iterative code using SB outperforms a recursive code using HM and obtains competitive results on a variety of platforms.
Springer
以上显示的是最相近的搜索结果。 查看全部搜索结果