Quote from CACM paper: cost of parallelism

["Ruth Ivimey-Cook" <ruth@xxxxxxxxxx>]

Folks,

Can the CSP world provide any counterexamples to the following? A colleage
wrote:

 ''I found this re-quote in the paper Optimistic Parallelism Requires
Abstractions, Kulkarni et. al., Communications of the ACM September 2009
Vol 52. No. 9, DOI:10.1145/1562164.1562188:

"Tim Sweeney, who designed the multithreaded Unreal game engine, estimates
that writing multithreaded code tripled software costs at Epic Games"

The original quote is from the webpage below, and has slightly different
emphasis:

"Implementing a multithreaded system requires two to three times the
development and testing effort of implementing a comparable
non-multithreaded system, so it's vital that developers focus on
self-contained systems that offer the highest effort-to-reward ratio."

Both the paper and the interview are worth reading, for what they say
about auto-parallelising and irregular application parallelism.''

Regards,

Ruth


http://www.anandtech.com/cpuchipsets/showdoc.aspx?I=2377