[concurrency-interest] Best CPUs for testing

Joe Bowbeer joe.bowbeer at gmail.com
Wed Dec 13 11:23:33 EST 2006


At the machine level, the DEC Alpha used to be the platform that would
expose the most bugs because it imposed the least order -- and was even
purported to flaunt this at every opportunity in order to shake out bugs.
(The principle of "least astonishment" turned on its head to benefit
testing.)

You can see in the cookbook that the Alpha has little respect for data
dependencies and consequentially needs a lot of memory barriers:

http://gee.cs.oswego.edu/dl/jmm/cookbook.html

On the other hand, since Alpha's Java VM implementation is so tightly
fenced, perhaps the Java apps running on it are *less* likely to do anything
astonishing? :-)


On 12/13/06, Péter Kovács <peter.kovacs at chemaxon.hu> wrote:
>
>
> A passage in "The Java Memory Model" section in "Concurrent Programming
> in Java" prompted me to ask the question: can the CPUs currently
> available on the market be ranked by their suitability for detecting
> concurrent programming problems.  [...]
>
> Any comment appreciated.
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20061213/55c41666/attachment.html 


More information about the Concurrency-interest mailing list