[Cuis] Curious drawing performance issue

[Dan Norton]

On 13 Aug 2015 at 16:01, Phil (list) wrote:

> On Thu, 2015-08-13 at 11:11 -0400, Dan Norton wrote:
> > On 13 Aug 2015 at 2:29, Phil (list) wrote:
> > 
> > > Dan,
> > > 
> > > 
> > > On Wed, 2015-08-12 at 22:17 -0400, Dan Norton wrote:
> > > > Hi Phil,
> > > > 
> > > > On Windows 7 with Cuis loaded and idle, Windows Task Manager
> >
> > > Performance reports 0% 
> > > > CPU Usage. Memory Usage is 1.28 GB.
> > > > 
> > > 
> > > What does the cpu usage show via World Menu -> Open... ->
> Process
> > > Browser for Morphic?
> > > 
> > 
> > It shows 12%.
> > 
> > > > With BouncingAtoms morphExtent: 500 at 450, stepTime 0, nAtoms
> 5000
> > > the report is 24 - 
> > > > 25% CPU Usage, 1.28 GB Memory Usage, 2 - 6 fps.
> > > > 
> > > 
> > > OK, so it's not platform specific poor performance.  If you
> close
> > > the
> > > BouncingAtomsMorph and wait for things to settle for a few
> seconds,
> > > what
> > > does Process Browser show the Morphic cpu usage as??
> > >
> > 
> > It shows 40%. However, Win7 shows 1 - 2%. Using Process Browser to
> measure absolute as 
> > opposed to relative CPU usage seems inaccurate to me. The
> /difference/ in Process Browser 
> > numbers while atoms is running vs not running agree with Win7: 1 -
> 2%.
> > 
> > I don't like for something to try to measure its own performance
> :-) - an independent 
> > instrument is preferable IMHO.
> 
> True, the external monitoring tools will tend to give you a better
> absolute picture as there may be some VM overhead (esp if plugins
> are
> involved) that Process Browser can't see.  However, CPU utilization
> as
> reported by the OS isn't terribly helpful unless you know how many
> cores
> the system has, was anything else active at the time, etc.
> ProcessBrowser, while limited, avoids all that since it shows what
> was
> used vs what was available to it. (usually 100/<num cores> % of
> total
> system CPU)  Just seemed to make for a better apples to apples
> comparison in this case to see if you are having the same issue
> re:
> Morphic chewing up cycles while otherwise 'idle'.
> 

Rummaging around in my system reveals:

Microsoft Windows 7 Home Premium
6.1.7601 Service Pack 1

Intel(R) Core(TM) i3-2125 CPU @
3.30 GHz

All Programs>Accessories>System Tools>Resource Monitor shows:

an appalling (to me, at least) >170 threads "if the only tool you have is a hammer, every 
problem looks like a nail" sitting there quietly for the most part. Resource Monitor shows four 
"CPUs" with CPU 1 and CPU 3 parked. Total CPU usage < 1%.

When Cuis is started, the Squeak.exe starts out with 17 threads, then eventually settles to 6 
threads. Total CPU usage < 1%.

When BouncingAtoms is running, set up as above, CPU 1 and CPU 3 remain parked, CPU 0 
< 5%, and CPU 2 > 95%. Occasionally something (GC?) cranks CPU 0 to 85 - 90% briefly.

When BouncingAtoms is stopped, all return to CPU 0 < 5%, CPU 2 <1%, CPU 1 and CPU 3 
parked.

So, the question is why should Cuis process browser report that Morphic UI is consuming 10 
- 14% ?

> >  
> > > > Cuis 4.2 2449, cogwin 15.22.3370
> > > > 
> > > >  - Dan
> > > > 
> > > 
> > > Thanks,
> > > Phil
> > > 
> > > > On 12 Aug 2015 at 17:44, Phil (list) wrote:
> > > > 
> > > > > I should also mention that the Morph window is ~ 500x450
> and
> > > > > #stepTime
> > > > > is set to 0 for anyone who wants to try to replicate...
> > > > > 
> > > > > On Wed, 2015-08-12 at 17:36 -0400, Phil (list) wrote:
> > > > > > I noticed a while back that something appeared to be going
> on
> > > with
> > > > > Cuis
> > > > > > drawing performance (at idle on my system Morphic
> consumes
> > > ~20% of
> > > > > VM
> > > > > > CPU *miniumum* according to ProcessBrowser) and this seems
> to
> > > give
> > > > > an
> > > > > > indication of what it's currently costing in drawing
> > > > > performance...
> > > > > > 
> > > > > > After seeing the Squeak 5.0 announcement, I was curious
> to
> > > see
> > > > > roughly
> > > > > > how much of a speed boost we might be able to expect
> from
> > > Spur
> > > > > down the
> > > > > > road.  So I decided to look at BouncingAtomsMorph to try
> to
> > > get a
> > > > > rough
> > > > > > apples-to-apples comparison and was quite surprised:
> Spur
> > > was
> > > > > faster,
> > > > > > but it was too much faster.  So I dropped back to Squeak
> 4.5
> > > and
> > > > > it also
> > > > > > performs much, much better than the Cuis version on the
> same
> > > VM.  
> > > > > Here
> > > > > > are the numbers I'm seeing using BouncingAtomsMorph with
> > > roughly
> > > > > > comparable (i.e. eyeballed) morph sizes and atom count
> set
> > > to
> > > > > 5000:
> > > > > > Squeak 5.0 (Spur VM from all-in-one download): 29-31 fps
> > > > > > Squeak 4.5 (Cog VM 15.25.3390): 24-26 fps
> > > > > > Cuis 2440 (Cog VM 15.25.3390): 6-8 fps
> > > > > > 
> > > > > > Granted BouncingAtomsMorph is not 100% identical from a
> > > source
> > > > > code
> > > > > > standpoint but it's not nearly different enough where
> I'd
> > > expect
> > > > > that
> > > > > > sort of difference.  Is this a platform-specific issue
> (I'm
> > > on
> > > > > Linux) or
> > > > > > are others noticing drawing issues as well?
> > > > > 
> > > > > 
> > > > > 
> > > > > _______________________________________________
> > > > > Cuis mailing list
> > > > > Cuis at jvuletich.org
> > > > > http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org
> > > > 
> > > > 
> > > > 
> > > > _______________________________________________
> > > > Cuis mailing list
> > > > Cuis at jvuletich.org
> > > > http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org
> > > 
> > > 
> > > 
> > > _______________________________________________
> > > Cuis mailing list
> > > Cuis at jvuletich.org
> > > http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org
> > 
> > 
> > 
> > _______________________________________________
> > Cuis mailing list
> > Cuis at jvuletich.org
> > http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org
> 
> 
> 
> _______________________________________________
> Cuis mailing list
> Cuis at jvuletich.org
> http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org