Benchmarking question

Ed Griffin edg at ll.mit.edu
Tue Jun 18 16:05:00 CDT 2002


Hi all, for many years I have stayed away from the quad processor servers 
due to the fact that the processors are slower than what you can get in a 
dual-processor box.  Has anyone tried any kind of benchmarking with say a 
dual 2.4GHz 2650 versus a quad 1.6GHz 6650?  (preferably with something 
that uses very large processes ie: >1GB and lots of floating point 
operations) I know the processors have different architectures and 
differing cache configurations, I am wondering if the hit in raw MHz you 
take on the Intel Xeon MP is overcome by the changes in architecture.  I 
have always managed to stay away from needing this much horsepower, but 
with things as they are my developers are always taking every CPU cycle 
available and things just keep getting bigger.  Any insight would be 
appreciated.  And by the way I do understand we are comparing "apples to 
oranges".

Thanks.

--Ed




*****************************************************************
Ed Griffin
Assistant Network/System Administrator
MIT Lincoln Laboratory - Weather Sensing
Phone:    (781) 981-2871
mailto:edg at ll.mit.edu
*****************************************************************




More information about the Linux-PowerEdge mailing list