[Macworld]

Post your comments for iMac performance evolves through the years here

[TiggerToo]

Fascinating, easy to see how we take the power we have these days for granted. my intel iMac with 4GB of ram, same size as the HD in that Grape iMac, powers through multiple tasks in a tenth of the time the old Powermac Dual G4 used to and I work in video production. Lot's of big files, lot's of compression and it doesn't even break a sweat.

[jmvr]

Why did you choose not to go the orher way, using iMacs with the system software they originally shipped with, same for the applications, to find out time needed to achieve determined goals ?
The original iMac (G3/233) with 256 Mb RAM and Mac OS 8.6 was a nice and fast machine, using the software available at the time.

[Wizardling]

Weird. I have an original Rev a bondi blue iMac upgraded to 384MB RAM (128 + 256) and a Rev c lime green iMac upgraded to 512MB (2*256) RAM. I'm not sure why you guys had trouble. Are you positive you used compatible RAM?

[aaronvegh]

The Camino test blows my mind! The notes say the test consists of opening five web pages on a local server. I can't fathom that taking 55 seconds on a modern iMac, never mind 40 minutes for the G3! I wish you'd included more information on what those pages consisted of.
I would also second the point about using contemporary software with each iMac: presumably we'd like to compare "best performance possible" for each system? While the Intel Macs were hampered by running in Rosetta, the earlier models were hamstrung running apps developed after they were discontinued.

[mariorelvas]

Yes, you are right. That would be a better way to measure how performance evolved.

Also, it would be interesting to see the evolution of the performance ratio between the entry level iMacs and their counterparts top of the line. Are we closing the gap?

Both of these comparisons would help us all answer the question: are we better served now?

[beegee225]

Yes... Use the contemporary software as a basis for each one. Then at least we're close to comparing apples to apples. Simpler operating systems had simpler tasks, not all the bells and whistles that come with today's programs. And, the work asked of them was simpler, too. You should be able to get that 333MHz grape one to matching 512mb easy enough. Answering the "better served now?" question is a good one.

[spinoza2]

I'm also not sure the point in artificially throttling down the Intel machine to run Rosetta, I don't see this being an accurate reflection of real-world performance comparisons. More realistic would be to use iLife 08 on the newest machine, as well as CS3 and Office 08, which are all universal binary. In other words, one should test the entire computing environment, and not just the hardware. This more accurately reflects the actual computing environment at any given moment in time.

[Hobeaux]

iLife 05? I mean c'mon. Yes it's a Universal Binary application, but so is iLife '08. If you're using the old app so that very old macs can be pitted against the new, then you really should have one additional category that pits the newest software against everything that will work with it.
tsk tsk. bad macworld!

[rcray]

Maybe it's finally time for me upgrade from my 400mhz G3 (blueberry) imac!

[XMattingly]

Sure, the hardware specs on iMacs have ramped up dramatically over the years, but does that mean they're truly better than last year's model?
I think a much more realistic test of performance would be running the operating system that came with each computer and software that was available at the time, rather than trying to shoehorn the latest and greatest on there.
10.3 with only 256mb of recognizable RAM? How do you expect to even be able to run any software on that kind of setup?

[whitedog]

Wow! Damned if you do and damned if you don't. For a long time people have been asking for comparisons between old and new Macs. Macworld finally delivers and they are raked over the coals for not setting up the tests right. All this proves is that you cannot really compare old and new - Macworld's efforts to do so were bound to fall short. Things have simply (or not so simply) changed too much. When the iMac came out, 56 k was considered a fast internet connection - and a built-in modem was a real luxury. At the time a 4 GB hard drive was huge. Where the iMacs lagged behind the competition was in RAM and video processing - and a slow 66 MHz system bus. But all they had to do was run OS 8.x and later OS 9.x., Internet Explorer, Outlook Express, Office 98, Pagemaker 6 and Photoshop 5. Besides the OS there were no Apple apps to speak of.

The evolution of the iMac between then and now does illustrate very nicely how computers in general have evolved. What hasn't changed much in ten years is the price. The first iMac in 1998, at $1,299, was $100 more than an entry level iMac is today. Factor in inflation and the difference is even greater.

You can compare specs; they have improved radically. What you cannot easily measure, however you structure the tests, is the increase in raw computing power. The apps that ran on early iMacs won't run on contemporary models. And the apps that run now won't work on vintage iMacs. Apples and oranges, no matter how you slice them. Backwards compatibility has its limits.

Clearly, as computers have gotten more powerful, our expectations have risen right along with them. As computers get more capable, the software we run on them grows exponentially more powerful as well. Nevertheless, with the exception of the switch from PPC to Intel processors and two changes in form factor, the changes over ten years have mostly been incremental and undramatic. A few GHz here, a few GB there; the decision to upgrade to a new computer was almost always a difficult one - and still is.

For those still using vintage iMacs perhaps the most remarkable thing is that the monitors have held up all this time. Nothing kills an old iMac more decisively than a darkening screen with the accompanying eye strain. Absent that hazard, people can plod along in OS X 10.3x almost indefinitely - and even in OS 9. Of course the web is almost dead to OS 9 - there are no modern browsers for it that can render current web content. Try navigating the Apple web site in Internet Explorer - it can't be done.

There are really only three questions to ask when trying to decide about getting a new computer: 1) Is there anything important I need to be able to do that the computer I have cannot handle? 2) Am I missing out on anything important by not upgrading? and 3) Can I afford it? It's a simple scale: two yes answers and you should get a new computer; two no answers and you can wait till one of those answers changes.

There is one other question people often ask for which there is no good answer: if I buy now will my new computer be rendered obsolete in a month or six when Apple upgrades its lineup? The only way to deal with this question is to answer the other three.

As Macworld's retro tests illustrate, there will always be a newer, faster, better iMac eventually.

[XMattingly]

Since hardware of former Macs weren't designed to run modern software, the only logical comparison would be to run tests on identical tasks on each machine, with the software and OS that was available for their time. The grape iMac was not only hobbled by running an OS several years beyond its heyday, but also with an anemic amount of RAM... all Macworld is doing is pointing out what we already know: if an old computer is even able to run the most modern OS, it'll be very slow. Duh. If you had that same grape iMac running OS 8 or 9, then tests like "open multiple folders" wouldn't have such a dramatic spread... I can guarantee you that grape iMac won't take 2:12 under its intended OS. I would wager that in some cases, you may even find the old machines outpacing the newer ones.

whitedog said:

the changes over ten years have mostly been incremental and undramatic.

From quarter to quarter that's mostly the case, but when you look at the long lens from the first to the latest, it's a very dramatic difference in hardware. Which makes comparing really old machines to new ones (both running modern software) bunk.

It was a valiant, and obviously laborious effort on Macworld's part. A pretty neat recap on what's been going on in the past decade, to boot. I would be very interested to see how the newest iMacs really stack up to the oldest ones performance-wise, for comparable tasks with their respective concurrent software & OS'.

[mariorelvas]

MacWorld could easily set up a meta test using their old reviews and lab tests of so called real-world computing ? boot time, opening apps, scrolling word, duplicate file, search and replace, and so on. MacWorld has all the information they need at hand. This was the first article http://www.macworld....imacreview.html but they have more.
This meta-test would show us how the iMac evolved in the real-world.
Just a suggestion to MacWorld that I would like to thank in advance.
The iMac Cometh