Size of 8700 GT

If the 8700 GT is merely an 8600 GT overclocked, doesn't that mean you can fit it in place of an 8600 GT?

 

You might as well overclock a 8600GT in the first place, nevermind replacing it with a more expensive part.

 

That would make sense.

As far as overclocking yourself, it sounds like a good plan but it also sounds dangerous if you don't know what you're doing. I don't think the average consumer wants to take that chance.

 

Laptop graphics cards are usually undervolted for a reason, and that is to avoid the excess heat generated when you overclock something. I would recommend finding out what the stock clocks for the 8700M GT are, what the manufacturer recommends, and other people who have the 8700M GT who have overclocked and posted their results online. I would figure out what seems safe, and put it as a goal if you really feel like overclocking.

 

its not literally an overclocked 8600m gt. functionally, spec wise, it is an overclocked 8600m gt. however, the 8700m gt is physically larger, mostly to accommodate the increase in thermal output.

by the same token, an 8600m gs is merely an overclocked 8400m gt. however, those two chips have a different thermal output and power draw. same story. the processor itself is the same, but there are certainly going to be little differences in the board the chip resides on. due to the increased heat of the 8700m gt, it is on the mxm IV interace. the 8600m gt is on the mxm III. if you could clock an 8600m gt to 8700m gt speeds, you would see exactly equal performance. but you can't get the 8600m gt to 8700 speed. im sure that in the factory the test out all the silicon. the 8600m gt's that can clock higher are attached to the larger boards and shipped off as 8700m gt. the ones that can't clock to 8700 speeds are branded 8600m gt - and then attached to the 8600m gt board.

 

Actually... while your explanation is fairly close, theres a few things off.
1. 8600GT is found on MXM II. 8700 on MXM III
2. An equally clocked 8600GT and 8700, would likely preform a little different due to modified ram configuration... Which way-I'm unsure it would make a great experiment.

but yeah... 8700s are a size up from the 8600s... Partially due to their heat, and the rest I suspect is their ram configuration. Not to mention it being essential to disperse its 50% increase in thermal output...

Quite frankly though-the gap between the 8600gt and 8700, are very small, significant, but small. Upgrading between them would be like upgrading between a 9600 and 9600 pro in a desktop in the ol days with ati... sure, you gained something-but it wasn't too thrilling.

 

Some of the 9500 pro's (most?) were unlockable, so you could get a full 9700 pro. So those were in fact more attractive. And the 8700M GT seems to outperform go7900gs, which outperforms 8600M GT (8700 GT> 7900 gs > 8600 GT).

But yeah, otherwise RogueThunder is correct.

 

ive seen a bencvh mark, not sure where from, but the 8700gt is bout 100 points lower in 3dmark 06 than a 7950gtx

 

3Dmark isn't all that reliable. Real-world performance (i.e. games) show that it's better than the 7900gs, but worse than the 7900gtx

 

fair enough, still quite impressive even thoguht it has half the fricken bandwith

 

The 8700 is optimisd for 3DMark, in the real games world it struggles to keep up with the 7950GTX, and is slightly worse than the 7900GTX, Compare the desktop 7950GT and 8600GT, exact same respective clock speeds and cores

 

i don't think 3dmark was meant as a tool to compare cards from different across different series.

 

Perhaps it wasn´t. What is pretty clear these days is that both manufacturers (Nvidia and ATI) found a way to achieve far superior 3dmark performances than their real in-game performances.

¿Deliberately misleading? I don´t know. I just know that I´m not getting the fps I want from theoretically powerful available cards. Everyone should be aware of this when considering a laptop for gaming purposes.

 

its hard to say that you get a "better" 3dmark performance than a real game performance, because 3dmark is only a benchmark for comparing graphics cards...

its sort of like comparing engine performance.

you have engine specs: like number of liters, horsepower, torque, etc.

those are like the specs of the gpu, but its hard to quantify real world performance from specs alone.

so we put this engine through its paces. 3500 EB points (engine bench). wow, what a really high score for EB points.

but it can only do 0-60 in 6 seconds? what the hell? how did my engine get such a HIGH EB points score compared to its 6 second 0-60 time?!?!?

you see the ridiculous nature of this argument? obviously with this series of engine, 0-60 in 6 seconds is just 3500 EB points.

it is what it is. its just a benchmark and the "performance" of the benchmark isn't defined in terms of real performance.

i hope that makes sense?