ATI spreading FUD ?

[Bjorn]

http://www.ati.com/products/PCIexpress/index.html

Other graphics companies have cards that are compatible with PCI Express, but they are still only AGP cards that are “bridgedâ€￾ by a second chip to be physically compatible with PCI Express slots on the motherboard. This architecture can only work at AGP speeds, and is more vulnerable to failure, performance bottlenecks and incompatibility with software applications.

According to NVidia, the bridged cards can work at twice the speed of the standard 8X agp. And why would a bridged card automatically be more vulnerable to failure ... ?

 

[IgnorancePersonified]

Dunno but more parts mean more chances of failure.

 

[AlphaWolf]

http://www.ati.com/products/PCIexpress/index.html

Other graphics companies have cards that are compatible with PCI Express, but they are still only AGP cards that are ?bridged? by a second chip to be physically compatible with PCI Express slots on the motherboard. This architecture can only work at AGP speeds, and is more vulnerable to failure, performance bottlenecks and incompatibility with software applications.

According to NVidia, the bridged cards can work at twice the speed of the standard 8X agp. And why would a bridged card automatically be more vulnerable to failure ... ?

I suppose that there is a heat sink on the bridge chip for a reason. Anytime you add a piece aren't you increasing the chance of failure? It's just one more thing that can go wrong.

 

[Nite_Hawk]

bjorn:

The speed issue they probably put there on the assumption that it's running at AGP 8x. They probably shouldn't focus so much on the "other companies" implementation and simply say that theirs utilizes the full spec. The failure comment I can buy though. Any time you add in extra complexity to a system (atleast where it could be done in a simpler way) you'll probably run a greater risk of failure.

I don't know if I'd exactly classify this as FUD, but I think it's kinda tacky to be badmouthing the competition.

Come on ATI, talk about how good your PCI-Express implementation is, not about how bad everyone elses bridge chips are!

Nite_Hawk

 

[IgnorancePersonified]

Maybe there talking about VidCard to Mem Transfers

 

[Bjorn]

I suppose that there is a heat sink on the bridge chip for a reason. Anytime you add a piece aren't you increasing the chance of failure? It's just one more thing that can go wrong.

The point is, do we know if the bridge chip will cause problems, more failures, incompatibility problems, only be able to run at AGP speeds ? I would say no.

 

[3dilettante]

PCI Express does handle traffic back to memory on its own lane. Even the so-called AGP 16X with the bridge chip to the core is going to have to worry about turnaround that a native implementation wouldn't.

 

[Bjorn]

Come on ATI, talk about how good your PCI-Express implementation is, not about how bad everyone elses bridge chips are!

Nite_Hawk

Agreed.

 

[AlphaWolf]

I suppose that there is a heat sink on the bridge chip for a reason. Anytime you add a piece aren't you increasing the chance of failure? It's just one more thing that can go wrong.

The point is, do we know if the bridge chip will cause problems, more failures, incompatibility problems, only be able to run at AGP speeds ? I would say no.

Well if you're going to nitpick ATI didnt say will, they said is more vulnerable to".

 

[Bjorn]

Well if you're going to nitpick ATI didnt say will, they said is more vulnerable to".

No, they didn't use the word will. But my point was that since they don't know if it's only going to be able to run at AGP speeds, be more vulnerable to failures then implying that i will is spreading FUD.

Maybe i should add that i don't really care about this. It's just that it's fun to see that Ati also have a PR/FUD department.

 

[gkar1]

You cared enough to make a thread about it
I want to see kyro.pdf type of proof before i call it fud

 

[Bjorn]

You cared enough to make a thread about it

Don't care as in don't think less of Ati because of it.

 

[Kombatant]

First thing that comes to mind is that a bridge chip, is..well.. a bridge, to translate AGP to PCI Express. I sincerely doubt that would enchance performance, since all it does is "connecting" the two interfaces. Your speed can only be as fast as your source is. And if your source is AGP, you don't get faster than than.

I plan to read some stuff on PCI Express (mainly to see what new brings to the table) one of these days, so I could be dead wrong with the above but it seems like a logical assumption.

 

[Bjorn]

First thing that comes to mind is that a bridge chip, is..well.. a bridge, to translate AGP to PCI Express. I sincerely doubt that would enchance performance, since all it does is "connecting" the two interfaces. Your speed can only be as fast as your source is. And if your source is AGP, you don't get faster than than.

The "source" is not AGP. It's PCI-Express.

 

[AlphaWolf]

First thing that comes to mind is that a bridge chip, is..well.. a bridge, to translate AGP to PCI Express. I sincerely doubt that would enchance performance, since all it does is "connecting" the two interfaces. Your speed can only be as fast as your source is. And if your source is AGP, you don't get faster than than.

The "source" is not AGP. It's PCI-Express.

Um no.

The graphics card will be sending out signals through its native agp, the bridge will be translating that to pci-express.

 

[Ardrid]

This should help clarify some things:

http://www.tomshardware.com/graphic/20040310/pcie-07.html
http://www.tomshardware.com/graphic/20040310/pcie-08.html

 

[Bjorn]

Um no.

The graphics card will be sending out signals through its native agp, the bridge will be translating that to pci-express.

I was thinking that the source is the motherboard. Cause the motherboard is the fixed point in this case which NVidia can't do anything about. The graphics chip is obviously AGP but a version that NVida says can run at AGP 16x speed.

 

[AlphaWolf]

Um no.

The graphics card will be sending out signals through its native agp, the bridge will be translating that to pci-express.

I was thinking that the source is the motherboard. Cause the motherboard is the fixed point in this case which NVidia can't do anything about.

Ok just a language thing maybe, but the source of most (99%?) of the data being transferred for the bridge chip will be the gpu.

 

[epicstruggle]

But my point was that since they don't know if it's only going to be able to run at AGP speeds, be more vulnerable to failures then implying that i will is spreading FUD.

How do you know ati didnt also research a bridge chip and are making assumptions based on their findings? Maybe thats why they can make certain claims, from first hand knowledge. Just my 2 cents.
epic

 

[Bjorn]

But my point was that since they don't know if it's only going to be able to run at AGP speeds, be more vulnerable to failures then implying that i will is spreading FUD.

How do you know ati didnt also research a bridge chip and are making assumptions based on their findings? Maybe thats why they can make certain claims, from first hand knowledge. Just my 2 cents.
epic

Ati's bridge chip <> NVidias bridge chip. They need Nvidias bridge chip and cards to have first hand knowledge.