Tue 25 Jun 2019 10:40 - 11:00 at 228AB - Parallelism and Super Computing II Chair(s): Sreepathi Pai

Irregular data structures, as exemplified with sparse matrices, have proved to be essential in modern computing. Numerous sparse formats have been investigated to improve the overall performance of Sparse Matrix-Vector multiply (SpMV). But in this work we propose instead to take a fundamentally different approach: to automatically build sets of regular sub-computations by mining for regular sub-regions in the irregular data structure.
Our approach leads to code that is specialized to the sparsity structure of the input matrix, but which does not need anymore any indirection array, thereby improving SIMD vectorizability. We particularly focus on small sparse structures (below 10M nonzeros), and demonstrate substantial performance improvements and compaction capabilities compared to a classical CSR implementation and Intel MKL IE's SpMV implementation, evaluating on 200+ different matrices from the SuiteSparse repository.

Tue 25 Jun

Displayed time zone: Tijuana, Baja California change

10:00 - 11:00
Parallelism and Super Computing IIPLDI Research Papers at 228AB
Chair(s): Sreepathi Pai University of Rochester
10:00
20m
Talk
Sparse Computation Data Dependence Simplification for Efficient Compiler-Generated Inspectors
PLDI Research Papers
Mahdi Soltan Mohammadi University of Arizona, Eddie C. Davis Boise State University, USA, Mary Hall University of Utah, Maryam Mehri Dehnavi University of Toronto, Payal Nandy University of Utah, USA, Catherine R. M. Olschanowsky Boise State University, USA, Anand Venkat University of Utah, Tomofumi Yuki , Kazem Cheshmi University of Toronto, Michelle Strout University of Arizona
Link to publication DOI Pre-print Media Attached
10:20
20m
Talk
Modular Divide-and-Conquer Parallelization of Nested Loops
PLDI Research Papers
Azadeh Farzan University of Toronto, Victor Nicolet University of Toronto
Media Attached
10:40
20m
Talk
Generating Piecewise-Regular Code from Irregular Structures
PLDI Research Papers
Travis Augustine Colorado State University, USA, Janarthanan Sarma Colorado State University, USA, Louis-Noël Pouchet Colorado State University, Gabriel Rodríguez Universidade da Coruña, Spain
Link to publication DOI