NVIDIA GeForce 8800GT Pictured

[Wildside]

double the bandwidth should not make any difference really since PCI-E right now doesn't max out the bandwidth of current cards...

i'm not a big expert, but there's very very little difference between agp 4x and 8x, and from what i hear PCI-E 8x isn't much different than 4x or 16x

well it's going from 2.5 GT/s to 5 GT/s. It sounds like it's a difference. I mean, we havent even seen a PCI-E 2.0 card on a PCI-E 2.0 mobo in benchmarks yet, the products r still in the making. Ur opinions may change, IMO.

 

[Sora]

well it's going from 2.5 GT/s to 5 GT/s. It sounds like it's a difference. I mean, we havent even seen a PCI-E 2.0 card on a PCI-E 2.0 mobo in benchmarks yet, the products r still in the making. Ur opinions may change, IMO.

We will see but once again...all it does is allow for more bandwidth to travel in the lane. Considering the 650i with x16 on the first and x8 on the second could SLI with pretty much the exact same if not BETTER than the 680i which has two full x16 lanes( using two 8800GTX's). If we haven't even maxed out the PCi-Ex16 lane to begin with, how is it going to help just making a new interface? An X1950XT on AGP compared to one on PCI-E are pretty much the exact same speed.

 

[mrfluffles]

just downclocked versions of the 8800gts

Wrong....

 

[Wildside]

We will see but once again...all it does is allow for more bandwidth to travel in the lane. Considering the 650i with x16 on the first and x8 on the second could SLI with pretty much the exact same if not BETTER than the 680i which has two full x16 lanes( using two 8800GTX's). If we haven't even maxed out the PCi-Ex16 lane to begin with, how is it going to help just making a new interface? An X1950XT on AGP compared to one on PCI-E are pretty much the exact same speed.

i see ur point. It sounds like u think PCI-E 2.0 wont be any difference from PCI-E 1.1, but i dont want to go against u though.

all i can really say is that we have to wait for benchmarks to see if what u say is right.

 

[dario03]

More bandwidth doesn't add performance unless you can take advantage of it. More bandwidth is like a bigger pipe but if you can push a ball through a small pipe with no problem than a bigger pipe won't get the ball through any quicker.

 

[maroon1]

fixed better!

where were u when i posted more info of the RV670???

Anyways, the 2950 (RV670), according to an article from theinquirer.net, it can outperform the R600:

ATI's RV670 can beat R600 - The INQUIRER

.

Thats according to the The INQUIRER

You need to show us this news from more reliable source

Anyway, HD2950pro have half the memory bandwidth of HD2900XT because of the the 256 bit memory interface. So, logically the HD2950pro should perform worse than HD2900XT at stock speeds

HD2950XT might be slightly better than HD2900XT because of the higher clock speed. But that video card is supposed to compete with 8800GTX/ultra not with 8800GT anyway

I think the RV670 will do better then the 8800GT. I mean, it's on a 55nm!! I

HD2900XT use 80nm process yet it consume more power and produce more heat than than 90nm Geforce 8800 series

 

[Wildside]

More bandwidth doesn't add performance unless you can take advantage of it. More bandwidth is like a bigger pipe but if you can push a ball through a small pipe with no problem than a bigger pipe won't get the ball through any quicker.

ya that is exactly what i was thinking of!! I was thinking all night for some reason on how much bandwidth makes a differ n that note of pipes is exactly on the analogy i was thinking.

ok u guys r right.

 

[SHaTTaFaKKaP]

looks so coool

 

[NosBoost300]

i didnt read this whole last page.. but scoring 10000+ stock makes this better than the gts.. the gts stock is only scoring 9000ish area... whats the deal here?

 

[Invisible]

i didnt read this whole last page.. but scoring 10000+ stock makes this better than the gts.. the gts stock is only scoring 9000ish area... whats the deal here?

better card? lol