Those benchmarks on that page are not factual. They are "leaked" numbers before the go5700 was even released! Here's a quote on their page.
"Obviously, it would be extremely hasty to jump to conclusions using a set of tests not run by ourselves regarding an as yet unannounced product.
However, assuming that these results are indeed accurate, it gives an interesting insight into what to expect from the GeForceFX Go 5700, as well as a very small teaser into what ATi's forthcoming M11 will offer. From these results, it seems that ATi's current dominance in the growing mobile market will continue unabated for the time being. Of course, only time will tell...."
I would even go as far as to say that those numbers are flat-out wrong. According to their page, they were testing, and I quote:
"The benchmarks which follow shows an ATi M10 chip at its lowest possible clock settings, against a GeForceFX 5700 Go overclocked with extra cooling. In other words, the worse case performance scenario for the ATi board against the best case scenario for nVidia's new contender, which obviously needs to be bore in mind when looking at these scores."
The M10 is the Mobility 9600....Not even the Mobility 9700, which at the time hadn't been released.
Those two things make the results real "fishy".
I think these benchmarks
are much more accurate. They tested the go5700 vs the Mobility Radeon 9700. The tests show that the M11, for the most part, outpreforms its counterpart, the go5700. HOWEVER
, the go5700 was tested with the "performance" settings enabled and not the "desktop replacement" settings which run 100 mhz (29%) faster.
Also, I may be nit-picking here, but the M11 was spotted a 6% CPU clockspeed advantage. (3.4 GHz vs 3.2 GHz)
But who knows, probably the M11 is better. If so, not by much. I am just trying to reassure myself that I didn't unknowingly buy an extremely inferior GPU.Order Update:
My 1712 has been shipped and is estimated to arrive this Tuesday! Awesome.