He said after login it takes 1-2 seconds quicker. Not actual boot of the Mac.
Again, he was referring to the login times. He never once mentioned the actual booting up of the Mac could be 1-2 seconds quicker, he said after login the load up time was quicker: "but for me i noticed that after the login the computer starts about 1-2 seconds faster."
I don't know where I've given you the impression that I was referring to boot times. I read the OP as referring to the time taken to log back in after the graphics system has been changed and that was what I was commenting upon.
Correct me if I'm wrong, but the allocation of RAM for the 9400M would have been done prior to actually logging into your user account. You bring up a good point about how the booting of the Mac could be quicker, but we are arguing about how quick the login is.
I'm reasonably certain that you are wrong on this one, I'm afraid. When a Mac is booted, a minimal driver set is loaded in order to get everything up and running in the fastest time possible. Since the log-in window is graphically undemanding, there's no need for any fancy 3d acceleration until much later in the game and the card only functions as a bare minimum VGA device. It's only after log-in that all the clever stuff happens.
Another point to back-up this view is that a change between cards only requires logging out and back in, rather than a restart.
He made an observation yet made no attempt to back up with any sort of logical evidence (i.e. like your post).
Sorry, this is an internet forum and not a peer-reviewed journal. His initial observation was:
like they said it doesnt make a difference when doing anything non-graphic intensive. but for me i noticed that after the login the computer starts about 1-2 seconds faster. i have tested this many times and it is true.
I've emphasised the important part. Is his assertion that "i (sic) have tested this many times and it is true." not good enough as evidence of a behaviour he has noticed?
Well, here is some evidence that should satisfy even you. I have run a series of timed tests using a stop-watch. Each test involved changing the graphics card and timing the interval between triggering the log-out event and the appearance of a reasonably large graphics file (around 650 MB) that had been added to the log-in items. Automatic log-in was enabled.
I have included combined errors of ± 0.3 seconds at the start of timing and +0.3/-0.0 at the end. I measured my reaction time to obtain this figure and the error bars are displayed on the graph. Also displayed on the graph is the mean lines for the restart times for the 2 systems.
You'll notice that there is a large patch of anomalous data starting from run 3. For some reason, my Mac hung at the log-out and needed a hard reset. It then took a few cycles to sort itself out. I've taken the liberty of removing this anomalous data and sorting the rest.
It's clear that there is a difference in initialisation times and that, on average, this difference is approximately 1 second. It's also clear that this difference (even allowing for errors) can be as high as 2.5 seconds. Interestingly enough, there is an overlap between the faster of the 9400 cycles and the slower ones from the 9600. It doesn't affect the overall result, though, especially when you take the error bars into consideration.
I argued that. It is a debate... is that so hard for you to understand?
You didn't engage in debate. You merely rejected his observation out of hand. Here's your post:
...
Not true. Things won't load faster just because you are using a dedicated GPU. Maybe in Snow Leopard, we'll see slight improvements..
I know what a debate is. We are having one now. It seems a little one sided from where I'm sitting, though. I've yet to see you present any relevant point.
And I believe he called me a tool, so he was the one who insulted me, not other way round (though that post seems to have been removed by a mod).
Sigh..
It appears to me that you fished long and hard for that insult, but then I'm just naturally suspicious...
