Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

raymondu999

macrumors 65816
Original poster
Feb 11, 2008
1,009
1
Hi all. Let's say, that the rumors are true, and the nVidia 9400M chipset is not 64-bit. But let's say I'm a 4GB person, or perhaps even 8GB (an nVidia PR has reportedly said that this is possible. Let's just assume that he's telling the truth)

So. The chip itself would be 64-bit, being the second gen of Penryn processors. Now, the chipset itself not being 64-bit, would that mean any performance hit? I'm not talking about "noticeable" performance hit. I'm talking about actual number crunching performance hit. Does anyone know whether the chipset itself not being 64-bit would hit performance?

Oh, and just to recap, this is assuming the new MBPs run fine with 4GB and 8GB as well, leaving RAM issues out of the pot. But as in the execution of 64-bit code and stuff. Thanks.
 
We are talking "theory" here. In real life most jobs are limited by how fast the user can think, move and click the mouse.

in theory, there is always one "bottle neck" that limits the speed the software can run at. What that bottle neck is may likely bounce around and be different second by second even within one application. A well designed computer system will have all of it's parts matched such that one part is always not the bottle neck.

So you ask if the width of the data path between the CPU and the GPU could be a problem. I imagine it would be the bottle neck that limits performance some of the time. But maybe it is the choke point for a millisecond every 10 to 15 seconds? Other times it's the head on the disk drive where the software has to wait for an 8 millisecond seek. Other times it might be the CPU.

It is the mark of a good system that EVERY part takes it's turn at being the choke point. If not then that part is to expensive and the budget should have been spent elsewhere. An engineer's goal if he is building bridges or computers is to design the system so that no one part is so expensive that it sucks budget away from other parts and forces those parts to be the weal link. A bridge designer wants the entire bridge to fail at the same time

OK back to your question: "Could this little part limit the performance of the entire system?" One hopes the answer is always "yes", no matte which part you ask about.
 
Ah... so the execution isn't slowed down, but the transferring of the code to be executed is slowed... or potentially slowed... and... are you THAT pissed at the glossy? (I'm referring to your last line there) :p
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.