9 Vs X performance compared.Does MacOs X take advantage of your system?

S

Samuel

Registered
Well I did the test last night. First, I downloaded a little program called Speedrun to test both systems: 9.2.1 & 10.1.5.

My setup: Pismo 400 wit two monitors, 192 mb ram, and 10 Gig hardrive.

The results:

10.1.5

Graphics 179
Hard Drive 185
Processor 138
Ram 254

Overall performance: 189

9.2.1

Graphics 510
Hard Drive 415
Processor 255
Ram 1362

Overall Prerformance: 636


Well, I just need to read the numbers...
Amazing !!
How is it possible? I supose it's the G3 processor, and that the difference will be smaller on G4 "MacOs X optimized computers"
This make me feel a little confused about Macos X. I would like to know how is it with Jaguar.
MacOs 10..1.5 is running just great on my pismo, but i didn't know (and noticed) that the difference in performance was so, so big.
It proves that MacOs X doesn't take advantage of my computer as well as Macos 9 does.
What do you think?
I'd like to see your test.

Thank's
 
i think thats a load of bs... os9 might be a bit slow sometimes on a prismo this i know but os9 sucks! crashing and old and such i would never take 9 over x even if it was on a 233 imac
 
OK My results with SpeedRun on a G3 B&W 350 with 348 Megs of Ram and a 20 gig HD:

JAGUAR:

Graphics: 185
Hard Drive: 175
Processor: 121
Ram: 230

Overall: 178



OS 9.2

Graphics: 402
Hard Drive: 328
Processor: 220
Ram: 1178

Overall: 532


Ok so at first glance I wanna give up trying to run OSX on my
NON-MacOs X optimized computer. I mean using by OSX, Im operating at 1/4 speed of OS 9.2......but...then I looked a little bit deeper.

This program gives you a chance to compare other systems with the same OS as you tested in, and I noticed that in both tests, my computer fit in exactly the same place in both tests: Right between an G3/300 and a G3/400.

So what to do? On all the tests on OSX the numbers were significantly smaller than on OS9. Does this mean this test is different for OSX than OS9 or does OSX hinder everything that much?

Ok just my input thanks
 
I believe this program is not useful/accurate for measuring how "fast" your computer is relative to other computers. I d/l both the OS 9 and X versions of this software and received radically different scores for each version.

Also, I seriously doubt any program which implies that OS 9.2 is anywhere from 3-9 times faster. That simply cannot be true, just from simple observation alone.

Anyway, don't get hung up on how "fast" your computer is relative to other users. Is it fast enough for you? The easiest way to stay up to date is to buy new Apple systems when they debut. Sell your old Mac, and apply the money you receive towards a new Mac. Just keep doing this everytime Apple puts out something new. That way you are only paying a fraction of the cost and you constantly have all new hardware.

If you don't want to do this, then you're probably going to suffer a performance decline at some point. You just can't have it both ways--fast performace AND no cost--it just isn't that way in the world of computers.
 
X may be slightly slower than 9, but the main difference is elsewhere: they manage multitasking differently !

I think the problem is not OS9 or X but the way SpeedRun is implemented. This software is very slow under X ! If it ever measure time the same way under both systems.
 
Those programs tick me off. They never give any indication of how those speeds actually affect my daily work. I haven't done a full battery of tests, but in my experience, OS X does everything as fast or faster except graphics manipulation (windows resize, grabber hand moves, scrolling, etc.). On Jaguar on my DP 1 gig (old model) graphics stuff is now acceptable if not blazing. I don't have any of the latest OS X games to analyze how they run, but Giants runs great in 10.1.5 (haven't tried in Jaguar yet). Network operations in OS X and disk I/O are much faster in OS X than 9 in my experience. Things like renders in various apps seem the same, though the dual processor config really opens up the computer for use that OS 9 could only dream of doing.
 
I don't have Mac OS lower than version 10.2.0 on my computer. How can I test the performance of OS 9?
 
Originally posted by mindbend
Those programs tick me off. They never give any indication of how those speeds actually affect my daily work. I haven't done a full battery of tests, but in my experience, OS X does everything as fast or faster except graphics manipulation (windows resize, grabber hand moves, scrolling, etc.). On Jaguar on my DP 1 gig (old model) graphics stuff is now acceptable if not blazing. I don't have any of the latest OS X games to analyze how they run, but Giants runs great in 10.1.5 (haven't tried in Jaguar yet). Network operations in OS X and disk I/O are much faster in OS X than 9 in my experience. Things like renders in various apps seem the same, though the dual processor config really opens up the computer for use that OS 9 could only dream of doing.
Click to expand...

I own the same dualie and I agree with mindbend. The only one major important difference that I have noticed is that, on FCP3, OS X really makes the app FLY! Heck, even on the Dualie 450 machine, it finally comes to life! Hurray for OS X!

Hey Digidesign peeps, if you are reading this... WHEN IS THE (I cannot use this f word here) OS X VERSION OF PROTOOLS COMING OUT? ARgghh... My audio editing system is still running OS 9.
 
Not a fair comparison since OS X has symmetric processing power. Meaning that the benchmark program doesn't get the whole CPU like it does in OS 9.
 
Originally posted by Ricky
Not a fair comparison since OS X has symmetric processing power. Meaning that the benchmark program doesn't get the whole CPU like it does in OS 9.
Click to expand...

Now it kinda make sense, but doesn't this means that if you only dedicate the Mac for one purpose, most of the CPU power is NOT utilised? Hmm... so what do you REALLY mean anyway?
 
I think the problem is not OS9 or X but the way SpeedRun is implemented.
Click to expand...

I agree totally. Surely the application will use a different set of libraries on each operating system, which means that you are effectively testing using two different applications. That is a large systematic error to start with.

I don't care whether OSX is x% slower or faster than OS9, the same as I don't care about the comparisons with PCs. The fact of the matter is that OSX allow ME to work faster.

R.
 
I'm with you. I don't think what this test shows is real. As I said, 10.1.5 is running great on my pismo except some graphic stuff. It makes me feel good that the differnce in performance on both operative systems isn't as big as this program says. I´m aslo very confortable with MacOsX :)
 
Originally posted by Ricky
Not a fair comparison since OS X has symmetric processing power. Meaning that the benchmark program doesn't get the whole CPU like it does in OS 9.
Click to expand...

Can you xplain this a little more?

Thanks:D
 
OS X shares the processor with other apps and processes, such as the Finder, for example. These processes take up a lot of memory and processing cycles to do. As you all probably know already, OS X takes a lot of processing power to run, but it also shares the processor cycles between programs, leaving the benchmarking test less power to work in, therefore giving a slower reading. OS 9 would allow the benchmark to take all of the cycles, giving it more processing power to work in.

Does this clear it up for you guys? :)
 
Originally posted by Ricky
OS X shares the processor with other apps and processes, such as the Finder, for example. These processes take up a lot of memory and processing cycles to do.
Click to expand...

He's right. Just run top in a terminal window and you should have about 50 or so processes running. I ran SpeedRun in Classic and it gave me a 603 overall and a 124 overall in OS X 10.1.5 beating out a 700 MHz eMac. Of course I just upgraded my Beige G3 to a 533 G4 but I don't think my 66 MHz bus could beat out an eMac. It also swung about 40% during an OS X retest for disk ratings. If the author wants to improve the program he should figure out how to let us adjust the nice rating for the program.
 
Originally posted by Ricky
OS X shares the processor with other apps and processes, such as the Finder, for example. These processes take up a lot of memory and processing cycles to do. As you all probably know already, OS X takes a lot of processing power to run, but it also shares the processor cycles between programs, leaving the benchmarking test less power to work in, therefore giving a slower reading. OS 9 would allow the benchmark to take all of the cycles, giving it more processing power to work in.

Does this clear it up for you guys? :)
Click to expand...

Thanks for the info. I understand it better now. Cheers.
 
You must log in or register to reply here.
Back
Top