I should warn everybody now that the PCIe 2.0 bridge is actually a PCIe 2.0 to PCIe 1.1 bridge.

What does this mean?

Well, it means your PCIe 2.0 card will basically be bottlenecked at PCIe 1.1 speeds when communicating to the CPU. False marketing from NVIDIA.

Here's a simple diagram of their "bridge":

. PCIe 1.1 upstream port to HOST
..........______|_____
.........|..................|
.........|...."Bridge"....|
.........|..................|
.........|___________|
.............| | | |
. PCIe 2.0 downstream ports to PCIe slots.


The only improvement that you may see with this is that communicating between other GPU's through the PCIe slot will be at 2.0 speeds (which is highly unlikely because NVIDIA uses their SLI bridge connector, not the slot).

AMD has TRUE PCIe 2.0 so EVERYTHING will be happening at 2.0 speeds.

I have read and confirmed this myself.

Right angle SATA ports FTL. And thanks for the info about the "PCIe 2.0," Nyte!

Well got to say nice looking board :)

So the only major difference is PCIe 2.0 support? The hell...

You don't need PCI-E 2.0 right now, the PCI-E 1.1 is not even maxed out yet, but I agree if this is false advertising by nVidia meaning your future PCI-E 2.0 upgrades will be bottle necked!!!!

I'm not gonna buy an nVidia board ever again, it's always gonna be Intel or AMD based from now on.

I want to see overclocking results, specifically with quad-core processors. I really hope some of the "bugs" they have fixed includes the down right poor quad-core overclocking.

by: Nyte

---

I should warn everybody now that the PCIe 2.0 bridge is actually a PCIe 2.0 to PCIe 1.1 bridge.

What does this mean?

Well, it means your PCIe 2.0 card will basically be bottlenecked at PCIe 1.1 speeds when communicating to the CPU. False marketing from NVIDIA.

Here's a simple diagram of their "bridge":

. PCIe 1.1 upstream port to HOST
..........______|_____
.........|..................|
.........|...."Bridge"....|
.........|..................|
.........|___________|
.............| | | |
. PCIe 2.0 downstream ports to PCIe slots.


The only improvement that you may see with this is that communicating between other GPU's through the PCIe slot will be at 2.0 speeds (which is highly unlikely because NVIDIA uses their SLI bridge connector, not the slot).

AMD has TRUE PCIe 2.0 so EVERYTHING will be happening at 2.0 speeds.

---

What does it matter, show me a card that actually uses PCI-E 2.0, then you will have a point about bottlenecking. By the time we see PCI-E cards that actually use PCI-E 2.0 we will be on PCI-E 3.0(or whatever it will be called). PCI-E 2.0 is nothing more than gimmicky hype, just like SATA II.

The main reason behind PCI-E 2.0 is for 8x multiple gpu systems and 1x/4x expansion cards. PCI-E 1.1 16x has more than plenty of bandwidth for any gpu, however 1.1 8x does not in some cases. With PCI-E 2.0 a x8 slot becomes the same as 1.1 16x in terms of throughput. With say the intel chipsets and their 16x 4x Crossfire setup, not having that bandwidth on the second card really killed performance in cases where you have no bridge. That is the only place 2.0 helps us.

On the server side of things 2.0 is way more important. Fewer lanes are needed for high bandwidth cards, which means more 4x/8x slots on server mobo's. This gives much more room for expandability, scalability, and longevity.

In terms of this bridge, it is mainly for marketing, and compatibility.

and this board will cost a bazillion dollars, yay......

by: Nyte

---

I should warn everybody now that the PCIe 2.0 bridge is actually a PCIe 2.0 to PCIe 1.1 bridge.

What does this mean?

Well, it means your PCIe 2.0 card will basically be bottlenecked at PCIe 1.1 speeds when communicating to the CPU. False marketing from NVIDIA.

Here's a simple diagram of their "bridge":

. PCIe 1.1 upstream port to HOST
..........______|_____
.........|..................|
.........|...."Bridge"....|
.........|..................|
.........|___________|
.............| | | |
. PCIe 2.0 downstream ports to PCIe slots.


The only improvement that you may see with this is that communicating between other GPU's through the PCIe slot will be at 2.0 speeds (which is highly unlikely because NVIDIA uses their SLI bridge connector, not the slot).

AMD has TRUE PCIe 2.0 so EVERYTHING will be happening at 2.0 speeds.

---

Nice, well explained but when you say everything will be happening at PCI-E 2.0 speeds you mean that 5GBit Bandwidth will be available and 150W be on tap, but as no card can use even the 2.5GBit bandwith of PCI-E 1.1 then there is little real time effect, damn the firsts graphics card to actually exceed the bandwidth offered by AGP 8x was the 8800GTX so we have a way to go yet before we can exceed 2.5GBit.
I just mention this just in case any of our less experienced members or visitors think that with a PCI-E 2.0 Gfx card they can plug it into a PCI-E 2.0 enabled slot on a motherboard and magically there is extra bandwidth/speed actually there, well it is there in so much as it's available, it is just not being used. Damn I'm confused now :eek: