Re: [PLUG] Use of dual core

[Antony P Joseph <antony@panathara.org>]

Hi

> Well, regardless of whether it's a hack or not, it does work and is a far 
> better approach than leaving the processor running at full speed even when it 
> isn't doing anything, which is why all modern CPUs do it.  Seriously, did I 
> not make it clear that there _isn't_ a performance hit already?  I don't wish 
> to be rude, but if you reply to someones message, do give them the bare 
> minimum courtesy by at least reading it first.
Probably this is due to the different understanding of the word
"performance" in computer industry. Different people interpret the word
different ways such as latency, response time, even throughput and so
on. This is what I used.

http://www.cs.wisc.edu/condor/tutorials/alliance98/slides/tutorial/sld004.htm

Performance suffers because there is a mechanism( an algorithm) involved
in determining the ideal frequency of the CPU(that itself may not be
guaranteed ideal frequency). If you plot a graph performance vs
throughput for a fixed amount of power consumption for different
governors, "powersave" will be highest throughput and smallest
performance , "performance" will be highest performance and smallest
throughput and "ondemand" may move along a curve between "performance"
and "powersave" depending on the algorithm used and characteristics of
the workload.

With regards
Antony

___________________________________________________________________________
Philadelphia Linux Users Group         --        http://www.phillylinux.org
Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce
General Discussion  --   http://lists.phillylinux.org/mailman/listinfo/plug