Originally posted by Programmer
There are two factors that Apple considers when choosing a GPU for the iMac -- what does it cost, and does it have the capabilities required to run their latest (or upcoming) hot software. CoreImage is a good example of the latter.
Well just how well is CoreImaging supported on current Mac hardware, that we don't really know. Yes we know that atleast a few PowerMacs and a couple of portables are "supported" but we don't know how well. As to the iMac well that most likely will be dealt with in the next rev, it is not unrealistic for people to expect very good performance with Core Imaging.
As observed up the thread, the target market for the iMac doesn't care which flavour of GPU is in the machine and frankly they shouldn't as long as it does the job.
That is a totally absurd suggestion. It is like saying that somebody going out and buying an automobile should not care what engine and transmission it has. It is a very poor investment if you buy a car with a manual transmission and you have no desire to drive one. There are whole host of other parameters that the average buy should be interested in when purchasing a car such as fuel economy.
People should care very much about what is inside the PC they purchase. That should not mena that they know and understand the machine to the depth of the average ARS user but they shouldn't be making the purchase as complete idiots either. It is reasonable to expect that people will want an iMac that supports the latest software from Apple even if they are not able to express what that software is. Much in the same manner as a person that wants an automatic transmission in his auto but has no idea how one works. Unfortunately we have no way to quantify performance (horse power & torque) of the GPU, so a reference to the hardware is a requirement. Lets not wander off into the world of benchmarking as I know that last statement leaves that open.
The problem with the ever chaniging nature of PC's is that it takes much more effort on the part of the average consumer to keep up with the changes in the technology. the choice between auto and manual and three different classes of engine horse powe for each automobile have been around for years. For the most part the consumer understands and accepts this arraingement. The options provide a way to solve the majority of potential cusotmer wants.
Heck I'd be happy to see Apple advertise their machines with two classes of GPU. Have one that offers "good" perfomance and another that offer 'maximum". AS long as Maximum actually delivers that and the machine with good performance offers a significant price differential the customers would be well taken care of. The problem is that Apple has had significant issue delivering maximum performance GPU's for some time now, if they continue to follow that pattern this idea would quickly become a joke.
The current "low end" GPU is actually a powerful unit compared to high end GPUs of 2-3 years ago, and it is perfectly capable of doing most tasks. Even compared to today's high end they are typically 33-66% the speed for most tasks, and for most users those high end GPUs are way faster than your typical iMac user needs anyhow.
How the current GPU's compare to processors of 2 or 3 years ago are of no concern to anyone. What should be of concenr is how the low end or even the high end of a few weeks ago compare to the recently arrived GPU's. The differential is rather huge between the new GPU.s and what was top of the line a few weeks ago. In effect the old low end stuff is so far behind as to not be significant anymore.
I think some of you guys have been staring at benchmarks too long, or are too focused on your own needs to realize that the iMac's target market has its own needs. Apple can't build a single machine for everybody, so they are aiming it as a particular class of user.
It is certianly true that Apple can't build a computer for everybody, but in the iMacs case they certainly have managed to build a computer for nobody. Don't get me wrong I like the appearance of the iMac it is the other short comings that turn me off personally and make it hard to reccomend. Is it possible for Apple to address the iMac issues and make a more marketable computer, sure. Does the market care - nope - at this point Apple will have to work real hard to sell the iMac3 beyond they debut surge they always get with new machinery. We can all hope that Apple has a winning formula.
Perhaps one day they'll decide the time is ripe to introduce a mini-tower or headless-iMac to address a different class of user, but that's not what the iMac is for.