[Elecraft] W9OY on P3

Fri Feb 19 18:41:04 EST 2010

I think there is a lot more important things in there other than 1 or 2
CPU's.  Architecture means quite a bit more.  How much L1/L2 cache, and
caching algorithms.  memory bandwidth, video memory bandwidth, which
ties in with FSB speeds.  Pipelining structures and how well each
application performs can vary depending on if its optimized for each
pipelining structure.  Not to mention that the OS can have more than its
fair share of play on how efficiently your applications even make it to
the CPU.  

I see Disk as less of an operation speed bottle neck with the price of
RAM today but as bad as Windows can get when its been on a computer for
a long time this can play a role.  Usually it has more to do with start
up and shut down times.  

When you're in Windows how bloated your registry has become and how many
TSR's you've loaded over the years can play to a large bit of difference
as well.  Thats why often just formatting a computer and reinstalling
the same version of windows for people usually makes them feel like they
just got a brand new computer.  

I'll be interested to see if I can get LP-Pan working on a box with a
Celeron processor in it... Its a 3.33Ghz box w/ 2GB of RAM but the lack
luster L2 cache that they put on the Celeron's can be a killer...  I may
have to see what that thing can take as far as real processors too... ;)

~Brett

On Fri, 2010-02-19 at 13:23 -1000, David Herring, K6DCH wrote:
> One point I forgot to make in my last post is that each point I  
> listed, both singularly and in combination with the others, directly  
> impacts CPU utilization.
> 
> 73
> David K6DCH
> 
> 
> Sent from my iPhone
> 
> On Feb 19, 2010, at 1:04 PM, "David Herring, K6DCH" <david at k6dch.com>  
> wrote:
> 
> > Hi All,
> >
> > I have a degree in computer science and 25 years in the industry, and
> > *I* wish someone could explain the disparity in performance among PC
> > makes and models. Hi Hi
> >
> > The fact is there are a very large number of factors that come into
> > play. I can list some here...
> >
> > CPU -- 1 or 2?
> >
> > Video card -- on board processing?  The same video board can display
> > vastly different performance depending on OS and driver.  There is
> > also differences between models within the same line.
> >
> > Memory -- not just how much you have, but how much is actually
> > available for allocation to the apps.
> >
> > Disk -- when you get to a point where the OS has to page memory out to
> > to disk, the speed of the disk will make a huge difference in
> > performance. Granted you are already compromised in performance at
> > this point, but a faster disk will minimize the pain. Sometimes by an
> > order of magnitude.
> >
> > Drivers -- version and vendor make a big difference here. Some drivers
> > perform well. Others are junk. Hard to know which is which.
> >
> > Junk Software -- many vendors load up the PC with pure garbage that
> > suck up resources and provide little benefit. Sadly you don't always
> > know that this stuff is running or what it's doing.
> >
> > Background processes -- antivirus, malware scanners, file indexing,
> > etc. could impact performance in non-intuitive ways.
> >
> > I doubt I've even scratched the surface here, but I think we begin to
> > see that the combinations of factors at play here makes for quite a
> > challenge in explaining the disparity among machines.
> >
> > Sometimes an older PC will run things faster because it has less
> > unneeded garbage on it and better drivers. It could also have less
> > bloat (XP is less bloated than Vista, for example).
> >
> > When I get a PC, the first thing I do is wipe the hard drive and load
> > my own clean OS. It's a pain but it mitigates many of the problems
> > here and I have always had better performance afterward.
> >
> > Now I do not just reload the OS from the vendor supplied disks -- that
> > just loads the same junk back on. I reload using a fresh copy
> > purchased "off the shelf."
> >
> > Hope that sheds at least a little light on the issue.
> >
> > 73
> > David. K6DCH.
> >
> >
> > Sent from my iPhone
> >
> > On Feb 19, 2010, at 12:14 PM, "Steve Ellington" <n4lq at carolina.rr.com>
> > wrote:
> >
> >> I wish SOMEONE could explain the disparities between CPU usage on
> >> various
> >> computers while running PowerSDR. I'm using a Dell 3.4Ghz CPU, 4GB
> >> ram. I'm
> >> running the same EMU-0202 at 96Ks/s. Windows XP, LP-Bridge
> >> etc.....My CPU
> >> usage hovers around 45%. I can stop PowerSDR and CPU usage drops to
> >> 4%.
> >> Obviously PowerSDR is eating up a lot of CPU time.
> >> My older Compaq with a 1.8Ghz processor really could not run PSDR at
> >> all.
> >> Then we hear stories about old clunky computers running PDSR  
> >> lightning
> >> fast...No one can seem to explain the difference.
> >> Mine runs fine but why would your CPU be less than mine when you are
> >> running
> >> more programs with less CPU and RAM?
> >> Steve
> >> N4LQ
> >> ----- Original Message -----
> >> From: "ab2tc" <ab2tc at arrl.net>
> >> To: <elecraft at mailman.qth.net>
> >> Sent: Friday, February 19, 2010 4:18 PM
> >> Subject: Re: [Elecraft] W9OY on P3
> >>
> >>
> >>>
> >>> I would have agreed if Windows had offered developers an easy way of
> >>> prioritizing threads and processes. But as far as I know it doesn't
> >>> (or
> >>> developers don't know how to use it). In my experience the
> >>> performance of
> >>> a
> >>> PC with 90% CPU load is miserable for all processes running on it.
> >>> With
> >>> that
> >>> said, I don't see why PowerSDR should incur that kind of CPU load
> >>> on a
> >>> 3GHz
> >>> machine. I am running XP home edition on a dual core Dell at 2.9GHz
> >>> and
> >>> 2Gb
> >>> of RAM. My CPU utilization is hovering between 15 and 30% with all
> >>> of the
> >>> following running:
> >>>
> >>> LP-Bridge
> >>> HRD
> >>> PowerSDR with EMU-0202 sound card at 192ks/s
> >>> VE7CC cluster client (highly recommended)
> >>> Iexplore composing this message
> >>> Thunderbird mail client
> >>> DX Atlas
> >>>
> >>> I can add more applications and the CPU barely nudges upwards. I
> >>> think
> >>> most
> >>> people would agree that a car that has to be driven always with the
> >>> accelerator nearly to the metal is underpowered and not much of a
> >>> joy. I
> >>> am
> >>> a firmware developer and we always worry whenever the CPU  
> >>> utilization
> >>> exceeds 50% even though we use OS's that allow intelligent
> >>> prioritization
> >>> of
> >>> tasks.
> >>>
> >>> AB2TC - Knut
> >>>
> >>>
> >>> Al Lorona wrote:
> >>>>
> >>>> Just a minor point: There might be a misconception that high CPU
> >>>> utilization means your computer is inadequate for the task.
> >>>>
> >>>> Actually, you want the CPU to work hard for you. It isn't only CPU
> >>>> you
> >>>> should worry about, it's what is called the 'run queue'. The run
> >>>> queue
> >>>> determines how long your job has to wait until it's serviced by the
> >>>> computer. It's okay to have 100% CPU (and in fact you want it) if
> >>>> you
> >>>> don't have to wait at all.
> >>>>
> >>>> A person assessing the performance of a computer looks at several
> >>>> other
> >>>> things besides CPU when determining what to tune for better
> >>>> performance.
> >>>>
> >>>>
> >>>> Don Wilhelm-4 wrote:
> >>>>>
> >>>>
> >>>>> I am using a 3.0 GHz Pentium 4 with 1 GB of RAM, running
> >>>>> WinXP Pro and the CPU utilization ranges from 50% to 90%, so  
> >>>>> anyone
> >>>>> thinking of choosing this alternative with a lesser computer had
> >>>>> better
> >>>>> think about a new computer first.
> >>>> <snip>
> >>>>
> >>>>
> >>>
> >>> -- 
> >>> View this message in context:
> >>> http://n2.nabble.com/W9OY-on-P3-tp4596769p4600120.html
> >>> Sent from the Elecraft mailing list archive at Nabble.com.
> >>> ______________________________________________________________
> >>> Elecraft mailing list
> >>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> >>> Help: http://mailman.qth.net/mmfaq.htm
> >>> Post: mailto:Elecraft at mailman.qth.net
> >>>
> >>> This list hosted by: http://www.qsl.net
> >>> Please help support this email list: http://www.qsl.net/donate.html
> >>
> >>
> >> ---
> >> ---
> >> ---
> >> ---
> >> --------------------------------------------------------------------
> >>
> >>
> >>
> >> No virus found in this incoming message.
> >> Checked by AVG - www.avg.com
> >> Version: 9.0.733 / Virus Database: 271.1.1/2697 - Release Date:
> >> 02/19/10
> >> 02:34:00
> >>
> >> ______________________________________________________________
> >> Elecraft mailing list
> >> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> >> Help: http://mailman.qth.net/mmfaq.htm
> >> Post: mailto:Elecraft at mailman.qth.net
> >>
> >> This list hosted by: http://www.qsl.net
> >> Please help support this email list: http://www.qsl.net/donate.html
> > ______________________________________________________________
> > Elecraft mailing list
> > Home: http://mailman.qth.net/mailman/listinfo/elecraft
> > Help: http://mailman.qth.net/mmfaq.htm
> > Post: mailto:Elecraft at mailman.qth.net
> >
> > This list hosted by: http://www.qsl.net
> > Please help support this email list: http://www.qsl.net/donate.html
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Elecraft at mailman.qth.net
> 
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html