Bang for the Buck, comparing Celerons to Xeons (and PIIIs?)

[Alan DuBoff]

David Welton wrote:

> Are you both doing the same make -j N ?  I think N in one case that
> was posted was 4, whereas I'm not sure there was even a -j on the
> initial post!
> 
> Make -j N defines how many concurrent processes may be running, which
> is obviously a number you can bump up if you are doing SMP.

No, I wasn't running more than 1 process, thats why George's real time is
lower than the user and sys combined, it's the sum of both processors. I get
my real time down by doing the same, but I don't have any tools that will show
me the processor activity. In regards to George's numbers, I can assume the
dual Celeron is actually faster than dual PIII 500s, it would be interesting
to compare with a dual PIII 550mhz box. There is probably no advantage with
the cache either way from the looks of it.

What are some of you other people using for monitoring the processors? I
thought that ntop would show multiple processors, but I think it only shows
network traffic...

In theory, there should be no advantage to running 4 processes on 2
processors, I'd like to see if that is true or not.

I was playing with the 2.3.18 code last night, there are a lot more warnings
and it takes substantially loger to compile.

-- 

Alan DuBoff
Software Orchestration, Inc.