Getting Started Documentation Glish Learn More Programming Contact Us
Version 1.9 Build 1556
News FAQ
Search Home

next up previous contents
Up: NOTE 232 - Parallelization Test Results Previous: Results

Subsections

Comments

Clark CLEAN Results

The Clark CLEAN tests show a modest improvement as we increase the number of processors. Previous Clark CLEAN deconvolution tests on the Origin 2000 have shown good speed-up using 32 processors. The benchmark reported here needs to be repeated using more processors.

The Linux cluster shows no improvement as we increase the number of processors! This is a perplexing result and needs some addition investigation.

We encounter AIPS++table locking problems on the Linux cluster at times; further investigation is needed to identify the problem. This was not a problem on the Origin.

IO Results

The roadrunner cluster is faster doing the parallel IO tests than the Origin with less than 32 processors.

The Origin show no improvement in performance past 32 processors on the IO tests when collecting Pablo statistics.

next up previous contents
Up: NOTE 232 - Parallelization Test Results Previous: Results   Contents
Please send questions or comments about AIPS++ to aips2-request@nrao.edu.
Copyright © 1995-2000 Associated Universities Inc., Washington, D.C.
Return to AIPS++ Home Page
2006-10-15