[JRD]

On my desk, I have a two computers:

G4 Cube
450 Mhz
512mb Ram
1 MB L2 Cache @ 225 Mhz
100 Mhz Bus
Rage 128 Pro (16 MB
Stock 20 GB Maxtor HD
10.3.3

12 Inch PowerBook (Rev B)
1Ghz
512mb Ram
512k Cache @ 1ghz
133 Mbz Bus
GeForce FX Go5200 (32 MB)
Stock 40 GB Toshiba HD
10.3.3

But the Cube "feels" snappier. Windows seem to move faster, applications launch and execute tasks quicker. The PowerBook, on the other hand, is sluggish... almost labored. It takes a second for sheets to decide they're going collapse back into a window and the same applications (DVD Studio Pro 2, for example) seem to take forver to launch.

So I ran XBench and compared the two. As expected, the PowerBook outranked the Cube in all the tests - in many cases by more than double - EXCEPT the hard drive. In that case the Cube had an overall rating of 63.35 and the PowerBook 56.68. The Random read/writes were especially divergent... an uncached write on the Cube was 70.17 vs. 47.88 on the Powerbook. The rest were between 10-15 points off.

I guess I'm looking for external confirmation that I'm interpreting these results correctly... a faster HD would make a significant difference.

[Michel_80]

Cube has 7200 rpm while your book has 4200. That is the main difference.

[tshow2k]

I have the exact same setup, except my Cube has a 40GB 7200 rpm drive. I agree that the cube does seem faster in almost every way so I ordered a 7200 rpm drive for my PB, and I'm eager to see the performance increase,

Greg

Originally posted by JRD:
On my desk, I have a two computers:

G4 Cube
450 Mhz
512mb Ram
1 MB L2 Cache @ 225 Mhz
100 Mhz Bus
Rage 128 Pro (16 MB
Stock 20 GB Maxtor HD
10.3.3

12 Inch PowerBook (Rev B)
1Ghz
512mb Ram
512k Cache @ 1ghz
133 Mbz Bus
GeForce FX Go5200 (32 MB)
Stock 40 GB Toshiba HD
10.3.3

But the Cube "feels" snappier. Windows seem to move faster, applications launch and execute tasks quicker. The PowerBook, on the other hand, is sluggish... almost labored. It takes a second for sheets to decide they're going collapse back into a window and the same applications (DVD Studio Pro 2, for example) seem to take forver to launch.

So I ran XBench and compared the two. As expected, the PowerBook outranked the Cube in all the tests - in many cases by more than double - EXCEPT the hard drive. In that case the Cube had an overall rating of 63.35 and the PowerBook 56.68. The Random read/writes were especially divergent... an uncached write on the Cube was 70.17 vs. 47.88 on the Powerbook. The rest were between 10-15 points off.

I guess I'm looking for external confirmation that I'm interpreting these results correctly... a faster HD would make a significant difference.

[WizOSX]

Do you guys have the energy settings on the PB set for highest performance? If not, then it's not a fair comparison between the Cube and PB.

[Lancer409]

was wondering about that .. post results after hdd swap and please let us know if it uses any more battery than usual. thanks a bunch

[iREZ]

If you could, try and bench it before and after the HD switch. I hate it when people say "it's soooo much faster, but when you see the facts its only a few points better (like 4 or 5) on Xbench. I really am tempted in upping my HD but voiding my Applecare worries me too much.

[JRD]

Originally posted by WizOSX:
Do you guys have the energy settings on the PB set for highest performance? If not, then it's not a fair comparison between the Cube and PB.

Yes, PB energy settings were highest performance and the unit was plugged in. I even went through the system settings and matched as many parameters as possible... same types of networks, sharing enabled, etc.

[arekkusu]

PB display is 12" AMLCD running at ~60Hz.

Cube display is ?????

If it is a CRT running at i.e. 100 Hz, that's part of your answer. Window dragging on PB = 60 fps. Window dragging on Cube = 100 fps.

[bpd115]

I don't see the logic..

HRZ do not equal FPS.....if so I'd be gaming on a CRT.

Nor does it explain the application launching speed etc.

[rag on a muffin]

network-playing Marathon on a powerbook 520c passive matrix display.

[arekkusu]

For the simple 2D operation of dragging a window, either of those two video cards can redraw the screen at several hundred fps. So Hz == fps in this case.

Other operations like app launching depend mostly on drive speed.

[dodo_nutter]

But even if Hz=FPS then the LCD is still giving out 60FPS. A TV is at 24-25FPS and things can move quickly on that cant they?

Surely the second or so wait JRD describes isn't anything to do with the refresh rate of the monitor or the FPS as if this was the case you would be talking abut the time between moving the mouse and the monitor getting updated which would be so fast no human could notice it. (like 1/60th of a second)

This has got to be a system problem with the OS not allowing the computer to perform to its full capacity OR the slow hard drive.

Please XBench before and after the new install and get back to us, nice to no how HDD upgrades effect performance (and battery life)

[Eug Wanker]

Originally posted by iREZ:
If you could, try and bench it before and after the HD switch. I hate it when people say "it's soooo much faster, but when you see the facts its only a few points better (like 4 or 5) on Xbench. I really am tempted in upping my HD but voiding my Applecare worries me too much.

Huh? Perception is probably more useful here than an overall Xbench score.

Remember, 4200 rpm laptop drives have a pretty good max transfer rate compared to 7200 rpm laptop drives. The key difference is randam read/writes, which would have a huge effect on perceived snappiness, but only a minor effect in the overall Xbench score.

[arekkusu]

JRD wrote:
But the Cube "feels" snappier. Windows seem to move faster...

This is the portion I was responding to.

[Cadaver]

Screen refresh rates has absolutely ZERO relationship with the speed (perceived or real) with which windows can move on the screen.

[arekkusu]

Wrong. Try this simple test:

1) Drag System Preferences window around on 15" Albook LCD. Result = 60 fps.
2) Drag same window around on connected CRT running at 1024x768, 120Hz. Result = 120 fps.

If the window dragging on the CRT doesn't look smoother to you than on the LCD, perhaps you should visit the optometrist.

[Megaquad]

Originally posted by arekkusu:
Wrong. Try this simple test:

1) Drag System Preferences window around on 15" Albook LCD. Result = 60 fps.
2) Drag same window around on connected CRT running at 1024x768, 120Hz. Result = 120 fps.

If the window dragging on the CRT doesn't look smoother to you than on the LCD, perhaps you should visit the optometrist.

Speed of dragging window or resizing it around has nothing to do with screen refresh, it has everything to do with speed of CPU and graphics.

Ability of graphics card to refresh screen 100 times per second doesn't mean graphic interface elements will refresh that many times, it depends on how much processor intensive is screen redraw or whatever interface is doing.

I wonder if 1.33 GHz 15" with 4200 rpm drive would also be slower compared to 7200 rpm machine on such trivial tasks...

[arekkusu]

The speed of any operation is limited by the slowest part of the computer. On a Rage 128 or any later video card, the slowest part of a 2D blit is the display (a LCD display will have a pixel refresh time between 16-25 ms. The CPU and video card can redraw the entire screen much faster than this.)

Don't wonder about it. TRY IT.

[Thorin]

OK, this is just my understanding of things, and I can't be bothered looking stuff up to back it up.

The slight ghosting you get on an LCD is down to it's response rate. The response rate is measured in milliseconds I believe, and the lower it is, the better the LCD is, and the less ghosting you will get.

The human eye can only see something in the region of 23.976fps. A person can not perceive the difference between 60fps and 120fps. Our eyes just aren't good enough .

Edit: Yeah... just done a quick search... the bit in italics is a load of rubbish, sorry! Turns out I didn't know what I was talking about at all

[V0ID]

Originally posted by dodo_nutter:
But even if Hz=FPS then the LCD is still giving out 60FPS. A TV is at 24-25FPS and things can move quickly on that cant they?

On a side note: Don't forget that on TV the video footage is blurred or (and?) interlaced which gives the impression of smooth movement on playback.

[velodev]

Originally posted by Michel_80:
Cube has 7200 rpm while your book has 4200. That is the main difference.

Maybe my Cube setup was different but the stock HD setup was not 7200.

But I doubt that... so I looked it up.

And yep, the 20gb HD was 5400. Not 7200.

On the other hand, the Cube was ATA66 and the Powerbook is ATA100. His problems shouldn't be the hard drive.

[freakboy2]

Originally posted by arekkusu:
Wrong. Try this simple test:

1) Drag System Preferences window around on 15" Albook LCD. Result = 60 fps.
2) Drag same window around on connected CRT running at 1024x768, 120Hz. Result = 120 fps.

If the window dragging on the CRT doesn't look smoother to you than on the LCD, perhaps you should visit the optometrist.

dude, you're wrong.. a lot of lcds have ghosting which is the rate at which the pixels lose their color.

otherwise, you're wrong. sorry.. the refresh rate is irrelevant to the speed at which windows open and drag around.

it's almost impossible for the human brain to perceive any difference in the motion of objects being refreshed past 60 hrtz.

i guess the easiest way to test this is to lower the refresh on a crt monitor to 60 and see if there is a difference in the speed of dragging windows around. I just did this and there isn't. sorry.

[JRD]

Originally posted by velodev:
Maybe my Cube setup was different but the stock HD setup was not 7200.

But I doubt that... so I looked it up.

And yep, the 20gb HD was 5400. Not 7200.

On the other hand, the Cube was ATA66 and the Powerbook is ATA100. His problems shouldn't be the hard drive.

So this is interesting. I'm clearly seeing a difference in hd performance from the benchmarking tests. What explains the gap?

[Megaquad]

It might be some kind of weird latency issues with connection to graphic card. Remember it's a portable and though Cube is older technology, laptop components often have some kind of sacrifices to ensure miniature dimensions.
Or it might be simply bad RAM?