The Problem with the Linpack Benchmark Matrix Generator
Jack Dongarra, Julien Langou
TL;DR
This paper investigates specific matrix sizes in the Linpack Benchmark where the matrix generator produces matrices with identical columns, highlighting potential issues in the benchmarking process.
Contribution
It identifies and characterizes the matrix sizes that lead to identical columns in the Linpack Benchmark matrix generator, revealing a flaw in the matrix generation process.
Findings
Certain matrix sizes produce identical columns in the generator
The flaw affects the reliability of Linpack Benchmark results
Guidelines for avoiding problematic matrix sizes are proposed
Abstract
We characterize the matrix sizes for which the Linpack Benchmark matrix generator constructs a matrix with identical columns.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsMatrix Theory and Algorithms