Will ATi jump on the SLI bandwagon?

What I see now is 2Q '05 ATi will release the R520 with 24 pipes (or so). Assuming nVidia has SLI up & running at that point, will we be seeing dual 6800 (or 6800 refresh) vs single R520?

OR.....

with Via & SiS also putting our SLI capable motherboards in competition of nVidia, will ATi release a SLI tech early next year so that we will see a nVidia SLI vs ATi SLI?
 
The talk is that RS480 and RS400 both have an extra 4 spare PCIe lanes in the Northbridge as well - what will ATI's partners do with those?
 
DaveBaumann said:
The talk is that RS480 and RS400 both have an extra 4 spare PCIe lanes in the Northbridge as well - what will ATI's partners do with those?
If they're smart and don't want to lose money, they'll do nothing.
 
Reverend said:
DaveBaumann said:
The talk is that RS480 and RS400 both have an extra 4 spare PCIe lanes in the Northbridge as well - what will ATI's partners do with those?
If they're smart and don't want to lose money, they'll do nothing.

eh if they are smart they will just put two x800pros on the same board and sell it for 800$ .
 
why would you want two X800s on a mohterboard? wouldnt the ability to have two PCIe slots on an ATI based mobo give the impression that ATI 'might' have a SLI solution down the line at some stage.

I dont see the point myself, but it could be usefull for surround gaming al a Matrox. Actually, with 1 vga on board, and 2 dual-head cards in the 2 PCIe slots, you get five possible monitors. I bet the flight sim guys will love that!

Ali
 
because you'd only loose 2 slots , the slot the card is and the slot of the cooler.

With sli u will loose at least 3 if not more .
 
From everthing ATI has said in the past, they consider SLI as a very small niche solution that is expensive, doesn't sell much, and really isn't profitable. It's only for bragging rights, and ATI has shown they focus on the big picture - making products to give them more marketshare, more units sold, and more profit. Making "concept cards" like SLI or MAXX just doesn't fit in with their current business model.
 
Bouncing Zabaglione Bros. said:
From everthing ATI has said in the past, they consider SLI as a very small niche solution that is expensive, doesn't sell much, and really isn't profitable.
Still, consider that right now they have no option but to say that.
 
Fodder said:
Bouncing Zabaglione Bros. said:
From everthing ATI has said in the past, they consider SLI as a very small niche solution that is expensive, doesn't sell much, and really isn't profitable.
Still, consider that right now they have no option but to say that.

But consider that it's still true. It's always been true with all SLI type solutions, especially in the home/retail market. If you consider that the top end cards are somewhat niche and sell in relatively small quantities (mid and low range OEM is where the real money is), how many people are going to want to pay for *two* high-end cards for SLI? It's an even smaller number. With PCs getting cheaper and graphics cards getting more expensive, we're already seeing VPUs costing a disproportionate amount of a PC - with SLI, you're looking at the VPU components costing significantly more than everything else put together

Why do you think there's only ever been SLI using dual cards in the home market once before? It's just not economic unless it costs virtually nothing to add to your card and so can be used for marketing/bragging rights - it's certainly not going to make any significant money in sales.

For me the main problem with SLI as a concept (leaving aside cost, heat, power, extra slots etc) is that when you are thinking about boosting your performance twelve months later, another 6800 isn't going to cut it. Anyone cutting-edge enough to want SLI willl be wanting the upcoming next generation card with the same processing power as SLI, plus all the new features of whatever the upcoming DirectX does that SLI on a year-old card won't provide.
 
Imagine a Maxx type card with 2 Rv410 on 0.09 process, maybe with a 256 bit bus. Could be nice card and still be a mainstream card if you think about costs next year.
 
DaveBaumann said:
The talk is that RS480 and RS400 both have an extra 4 spare PCIe lanes in the Northbridge as well - what will ATI's partners do with those?

Some 1X PCIE on the designs I've seen.
 
DaveBaumann said:
The talk is that RS480 and RS400 both have an extra 4 spare PCIe lanes in the Northbridge as well - what will ATI's partners do with those?

Ok, IF ATi partners make a dual PCIe ATi-based motherboard (20 total, 16 & 4 lanes), doesn't that pretty much force the market in that direction? SLI will then be on nVidia, Via, SiS, & ATi motherboards. Nothing pushes markets faster than a "standard", right?
 
How much do you guys think Nvidia has sunk into the R&D and marketing for it's SLI solution? Do you think it was a bad investment - i.e do you agree with ATI? I already know jvd, BZB and martrox do so everyone else please chime in with your thoughts on how sound a business/marketing strategy Nvidia has with SLI :D
 
It seems to me that with the development of newer graphics chips taking so long to come to market & costing so much to develop (compared to the past), that an SLI setup is a good way to increase performance without having to put so much money into development.
 
trinibwoy said:
How much do you guys think Nvidia has sunk into the R&D and marketing for it's SLI solution? Do you think it was a bad investment
Yes and no. Yes from a technical standpoint since it's overkill from hell right now, but no from a marketing standpoint as it gives 'em absolutely brilliant and clearcut bragging rights.
 
jvd said:
With sli u will loose at least 3 if not more .

Not necessarily with single lot cards as you can fit numerous boards between. It'll be more of an issue if the power from the second PEG x16 slot is diverted from elsewhere.
 
DaveBaumann said:
jvd said:
With sli u will loose at least 3 if not more .

Not necessarily with single lot cards as you can fit numerous boards between. It'll be more of an issue if the power from the second PEG x16 slot is diverted from elsewhere.

What would be the smallest PEG multiplier that would be good for the second card then?
 
Back
Top