Welcome to the MacNN Forums.

If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below.

You are here: MacNN Forums > Hardware - Troubleshooting and Discussion > Mac Desktops > Window resizing on the G5. What is the bottleneck?

Window resizing on the G5. What is the bottleneck?
Thread Tools
Eug
Clinically Insane
Join Date: Dec 2000
Location: Caught in a web of deceit.
Status: Offline
Reply With Quote
Nov 11, 2003, 01:26 PM
 
All I can say is that the Dual 2.0 with Radeon 9600 seems faster (resizing the same window) than the Single 1.8 with GeForceFX 5200.

What is the bottleneck? The video card and AGP bus? The CPU MHz? Memory bandwidth is the same on both machines.
     
Administrator
Join Date: May 2000
Location: California
Status: Offline
Reply With Quote
Nov 11, 2003, 04:27 PM
 
Move the 9600 over to the 1.8 and retest to determine if it's the vid card. If it isn't, move a 2 GHz CPU module over to the 1.8 and retest. If you have both machines in front of you, it should not take too long to narrow it down.
     
Eug  (op)
Clinically Insane
Join Date: Dec 2000
Location: Caught in a web of deceit.
Status: Offline
Reply With Quote
Nov 11, 2003, 04:34 PM
 
Originally posted by reader50:
Move the 9600 over to the 1.8 and retest to determine if it's the vid card. If it isn't, move a 2 GHz CPU module over to the 1.8 and retest. If you have both machines in front of you, it should not take too long to narrow it down.
Thanks for the non-answer.

I'm just wondering if anyone has compared a 1.8 side by side with a 1.8, with a 5200 vs. a 9600. I suspect it's mainly CPU speed.

Actually, I'm wondering if anyone knows the theory behind it.
     
Professional Poster
Join Date: Mar 2002
Location: Boston
Status: Offline
Reply With Quote
Nov 11, 2003, 05:44 PM
 
Maybe I'm missing something (which is normal) but why is that a non-answer. It seems very logical to swap cards and see if there's a speed difference.

Mike
     
Mac Elite
Join Date: Jul 2001
Status: Offline
Reply With Quote
Nov 11, 2003, 05:50 PM
 
It WAS an answer, just not to the question posed.
Actual conversation between UCLA and Stanford during a login on early Internet - U: I'm going to type an L! Did you get an L? S: I got one-one-four. L! U:Did you get the O? S: One-one-seven. U: <types G> S: The computer just crashed.
     
Eug  (op)
Clinically Insane
Join Date: Dec 2000
Location: Caught in a web of deceit.
Status: Offline
Reply With Quote
Nov 11, 2003, 07:17 PM
 
I took it as a joke, considering he said to swap CPUs. Anyways, it's moot, since I was just testing it in the store.

Has anyone compared two 1.8s side by side, with different video cards? It seems odd that only a dual 2.0 with Radeon 9600 is fast enough to give (relatively) non-jumpy window resizing.
     
Administrator
Join Date: May 2000
Location: California
Status: Offline
Reply With Quote
Nov 11, 2003, 08:01 PM
 
Eug, your post implied that you had both G5s in front of you. I assumed on your desk, store units didn't come to mind. I certainly do not have any G5s in front of me for testing.

While the CPU modules do unplug, on second thought it seems simpler to disable the 2nd CPU in software on the G5. The motherboards are apparently identical except for the missing 2nd CPU socket, so you would be testing a straight 200 MHz increase. But the store people might not appreciate CPU or video card swapping for benchmarking.

ps - the early G5 reports described physically unplugging the 2nd CPU for certain tests. The "independent lab tests" used this procedure if I recall correctly. So the CPU modules are removable. I was assuming that a 2 GHz module plugged into a 1.8 motherboard would still operate at 2 GHz.
     
Posting Junkie
Join Date: Jun 2003
Location: Dangling something in the water… of the Arabian Sea
Status: Offline
Reply With Quote
Nov 11, 2003, 09:14 PM
 
Originally posted by reader50:
Eug, your post implied that you had both G5s in front of you. I assumed on your desk, store units didn't come to mind. I certainly do not have any G5s in front of me for testing.

While the CPU modules do unplug, on second thought it seems simpler to disable the 2nd CPU in software on the G5. The motherboards are apparently identical except for the missing 2nd CPU socket, so you would be testing a straight 200 MHz increase. But the store people might not appreciate CPU or video card swapping for benchmarking.

ps - the early G5 reports described physically unplugging the 2nd CPU for certain tests. The "independent lab tests" used this procedure if I recall correctly. So the CPU modules are removable. I was assuming that a 2 GHz module plugged into a 1.8 motherboard would still operate at 2 GHz.
You can't plug a second CPU into the 1.8 mobo. There is no socket.
     
Professional Poster
Join Date: Dec 2002
Location: Partying down with the Ewoks, after I nuked the Death Star!
Status: Offline
Reply With Quote
Nov 11, 2003, 09:45 PM
 
Here is my take on the matter.

1) expose is really doing the same thing, actually more complicated and it is smooth as silk on a Dual G5. So why is window resizing slow?

2) Ars Techinica ask the same question:
http://www.arstechnica.com/reviews/0...wm-performance

3) I think it is just software optimization that needs some work.

4) I almost NEVER resize windows manually anymore, that is really what the green button is for. When I do I am not losing any sleep over a little stutter.

"Hello, what have we here?
     
Mac Enthusiast
Join Date: Jan 2003
Status: Offline
Reply With Quote
Nov 12, 2003, 03:47 PM
 
hmm, maye it's because the 9600 blows the 5200 out of the water?
Midshipman 3/C, USNR
     
Mac Enthusiast
Join Date: Mar 2002
Status: Offline
Reply With Quote
Nov 12, 2003, 04:55 PM
 
I'm not sure about the comparison to Expose. Each window is basically being mapped as a texture by Quartz extreme (as I understand it). Expose just scales that texture using the GPU to great effect no doubt. Window resizing requires the application to re-layout its contents based on the instantaneous window size.

Still, my three year old Pentium III laptop at work was quite a bit smoother at window resizing than the G5 I played with at the Apple Store.

Originally posted by Landos Mustache:
Here is my take on the matter.

1) expose is really doing the same thing, actually more complicated and it is smooth as silk on a Dual G5. So why is window resizing slow?

2) Ars Techinica ask the same question:
http://www.arstechnica.com/reviews/0...wm-performance

3) I think it is just software optimization that needs some work.

4) I almost NEVER resize windows manually anymore, that is really what the green button is for. When I do I am not losing any sleep over a little stutter.
     
Professional Poster
Join Date: Dec 2002
Location: Partying down with the Ewoks, after I nuked the Death Star!
Status: Offline
Reply With Quote
Nov 12, 2003, 07:01 PM
 
Originally posted by Boochie:
Still, my three year old Pentium III laptop at work was quite a bit smoother at window resizing than the G5 I played with at the Apple Store.
Ya but remember the interface is different, OSX looks much better.

"Hello, what have we here?
     
Dedicated MacNNer
Join Date: Jun 2001
Location: St Paul, MN
Status: Offline
Reply With Quote
Nov 12, 2003, 07:12 PM
 
Originally posted by Landos Mustache:
Ya but remember the interface is different, OSX looks much better.
More important that looks, OS X's Quartz interface allows for such fantastic things as Exposé and the picture resizing in iPhoto, to name a few. I'll take these things over fast-window resizing any day, although I wouldn't complain if they could figure out a way to make window resizing in OS X not suck
-Cory Bauer
cbauer@mac.com
http://www.sboobtv.com
     
Mac Enthusiast
Join Date: Mar 2002
Status: Offline
Reply With Quote
Nov 12, 2003, 09:18 PM
 
Agreed. It shouldn't take a $3000 top-of-the-line desktop machine to get smooth window resizing.

Originally posted by Cory Bauer:
More important that looks, OS X's Quartz interface allows for such fantastic things as Exposé and the picture resizing in iPhoto, to name a few. I'll take these things over fast-window resizing any day, although I wouldn't complain if they could figure out a way to make window resizing in OS X not suck
     
Mac Elite
Join Date: Jan 2003
Status: Offline
Reply With Quote
Nov 12, 2003, 09:38 PM
 
Originally posted by Mister Elf:
hmm, maye it's because the 9600 blows the 5200 out of the water?
Bingo. The FX5200 is a hideous abortion of a graphics card. On the other hand, the 9600 is quite a nice piece of hardware.


"Do I need to draw a diagram for you then to tell you that nerdy 16-17 year olds, fat chicks and old men turn my crank then? Will you understand it then or don't you follow still chris." - Landos Mustache
     
   
Thread Tools
Forum Links
Forum Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On
Top
Privacy Policy
All times are GMT -5. The time now is 07:26 PM.
All contents of these forums © 1995-2011 MacNN. All rights reserved.
Branding + Design: www.gesamtbild.com
vBulletin v.3.8.7 © 2000-2011, Jelsoft Enterprises Ltd., Content Relevant URLs by vBSEO 3.3.2