ATI's SLI to work without connector

[philler]

I have been waiting for this a long time..

ATI's wayof doing SLI called AMR (ATI Multi Rendering) will work differently from Nvidia's now world famous SLI. ATI will make its SLI work without a small printed circuit board (PCB) to interconnect the cards.
SLI uses two PCIe graphic ports and in the case of Nforce 4, works at PCIe 8X for each card and renders your picture with both cards. For this operation Nvidia SLI needs a special interconnection between two graphic cards. It?s a small cable or PCB that transfers up to four GB per second from one card to another. For this operation, coordination is crucial as one card needs to know what the other card is doing.

ATI found a way to make its SLI work without this interconnection. In ATI's case, whenever it makes its SLI ready, you will just plug two graphic cards in motherboard powered with ATI chipset and it will all work without interconnection PCB or cable.

The upcoming chipset will have hardware support for dual graphic card rendering mode. It will be all done through the chipset and we guess that it will be wired in the new RS482 and RX482 chipsets. All the communication between two cards will be done through motherboard or chipset wires.

ATI is working hard to catch up with Nvidia in the SLI game, and it will try to make its chipset and drivers ready for launch as soon as possible. This is expected sometime in Q2 2005. µ

 

[MegaWorks]

Let me guess, you work for ATI?

 

[philler]

Nope...i'm just rather fond of ATI .. because ATI has better picture than Nvidia's cards.. and the new solution without the PCB cable is just fantastic... i recently update my knowlegde of ATI's upcoming hardware.. and at last my wishes were fullfilled .. i just thought that you guys would like to get your knowlegde extended...

 

[Appledrop]

:/, my geforce 2 ultra seems to look better than my x800xtpe.

 

[philler]

what.. Geforce 2???..

 

[otispunkmeyer]

u just cut n pasted that off the inquirer no?

 

[mircea]

Well I don't think this is really that great. I mean, on Nvidia you have letancies because of 2 inces of circuit extension to have the GPU talk to each other.
Now ATI is doing it all trough the motherboard, which will probably take longer. So why is it so great not to have the PCB? Plus anyone that wanted to play with ATI but already got a MOB now they have to look for a new MOB. What's so great about that?
Nvidia allowd any chipset to go SLI not only the NV4. VIA and SIS are making their chipsets SLI capable. Now if you want to go Dual card with ATI you have to go with new cards, new MOB and un yet umproven shipset.

 

[dguy6789]

The ATI chipset overall is faster than the nforce 4. Going through the motherboard means it will be faster, not slower, as nvida's sli is limited to the 4GB/s communication between the two boards, while the AMR will be able to use the chipset, which can be much faster.

 

[chilled]

Originally posted by: otispunkmeyer
u just cut n pasted that off the inquirer no?

He did indeed!

 

[mircea]

Originally posted by: dguy6789
The ATI chipset overall is faster than the nforce 4. Going through the motherboard means it will be faster, not slower, as nvida's sli is limited to the 4GB/s communication between the two boards, while the AMR will be able to use the chipset, which can be much faster.

LINK please.
I'm no Nvidia fanboy (I can't afford to be a fan of anyone) but I have yet to see the ATI chipset working side by side with NF4 and then to see them in Dual GPU. After that we can talk. Until then, you presume optimistic, I presume pesimistic.

 

[Sunner]

Originally posted by: dguy6789
The ATI chipset overall is faster than the nforce 4. Going through the motherboard means it will be faster, not slower, as nvida's sli is limited to the 4GB/s communication between the two boards, while the AMR will be able to use the chipset, which can be much faster.

How exactly would that be faster, seeing as they'll have to work across the PCIe bus, which will be 2 GB/Sec bidirectional?

 

[biostud]

Bah, why can't the make a SLI/AMR that works on all PCIe boards no matter which chipset it uses.

 

[Stern]

Originally posted by: mircea

Originally posted by: dguy6789
The ATI chipset overall is faster than the nforce 4. Going through the motherboard means it will be faster, not slower, as nvida's sli is limited to the 4GB/s communication between the two boards, while the AMR will be able to use the chipset, which can be much faster.

LINK please.
I'm no Nvidia fanboy (I can't afford to be a fan of anyone) but I have yet to see the ATI chipset working side by side with NF4 and then to see them in Dual GPU. After that we can talk. Until then, you presume optimistic, I presume pesimistic.

yesyes, link?

 

[dug777]

Bah, why can't they make a SLI set-up affordable

 

[dguy6789]

You guys are members of the ANANDTECH forums, you think you would keep up to date with articles ANANDTECH has posted fairly recently at that. Just look in the chipsets section, and they have a preview on the ati chipset, along with the integrated graphics on it, and it is very close the the nf4, but usually wins by a few percentages.

 

[deveraux]

Originally posted by: mircea

Originally posted by: dguy6789
The ATI chipset overall is faster than the nforce 4. Going through the motherboard means it will be faster, not slower, as nvida's sli is limited to the 4GB/s communication between the two boards, while the AMR will be able to use the chipset, which can be much faster.

LINK please.
I'm no Nvidia fanboy (I can't afford to be a fan of anyone) but I have yet to see the ATI chipset working side by side with NF4 and then to see them in Dual GPU. After that we can talk. Until then, you presume optimistic, I presume pesimistic.

I'm no fanboy either and I think the reason (speculation here) that the chipset would be faster is that with nVidia's SLI, each GPU has to know what the other is doing hence adds overhead. While, if its done in the chipset, then the individual GPUs will get their own set of data to render without the need of knowing what the other is doing. Of course, if the chipset itself isn't fast enough or efficient enough, then all of this won't matter.

 

[philler]

yes i copied and pasted it from the inquirer.. because i have been talking alot of the new chipset from ATI, and some have asked me "does that really exist.. Etc.".. and then i have made this topic to answer their question.. ain't that allowed??.. just giving noobs some information..

 

[philler]

you got a point...perhaps.... well see.. as soon the mobo's on the market anandtech and tomshardwareguide Etc... will start testing them... well see..

 

[DaveSimmons]

Originally posted by: philler
yes i copied and pasted it from the inquirer.. because i have been talking alot of the new chipset from ATI, and some have asked me "does that really exist.. Etc.".. and then i have made this topic to answer their question.. ain't that allowed??.. just giving noobs some information..

It's allowed, but you are supposed to list your source when quoting someone, and for bonus points provide a link.

ATI found a way to make its SLI work without this interconnection. In ATI's case, whenever it makes its SLI ready, you will just plug two graphic cards in motherboard powered with ATI chipset and it will all work without interconnection PCB or cable.

This is going to hurt ATI. Many people want intel chipsets for their P4s, and nvidia chipsets for their A64s.

Of course right now you can't even buy most ATI cards, so nv SLI is getting a long head start.

 

[philler]

okay.. that makes sense to me.. sorry.. ain't gonna happend again..

 

[otispunkmeyer]

Originally posted by: Sunner

Originally posted by: dguy6789
The ATI chipset overall is faster than the nforce 4. Going through the motherboard means it will be faster, not slower, as nvida's sli is limited to the 4GB/s communication between the two boards, while the AMR will be able to use the chipset, which can be much faster.

How exactly would that be faster, seeing as they'll have to work across the PCIe bus, which will be 2 GB/Sec bidirectional?

good call!