Will ATi jump on the SLI bandwagon?

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 - 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

I think you have to ask yourself if the marketing you can do with SLI is worth the cost of R&D. You won't get money back on the sales, but you may decide that the cost is worthwhile as a form of advertising that helps raise awareness of your brand.

In this case I think it's definately a marketing led product, but once again seems to show that Nvidia have taken their eye off the OEM ball, as SLI is not an OEM friendly product.
 
As someone who doesn't upgrade a video card every year I can certainly see the benefits of SLI. Having said that if mobos with more than 16 lanes on the northbridge cost an arm and a leg or if there are problems introduced by SLI then it might just fizzle.

I want ATi to come up with their own "SLI" system, especially (as I understand it) it wouldn't require extensive R&D since from the R300 they are multi-core enabled, and X800's grouping of pipelines into quads.

And if product cycles are getting longer and longer, SLI could help combat that somewhat, allowing people instant doubling (or whatever) of performance for a low(er) cost. In the end, it's more choices for the consumer and I always think that is a good idea.
 
Bouncing Zabaglione Bros. said:
I think you have to ask yourself if the marketing you can do with SLI is worth the cost of R&D. You won't get money back on the sales, but you may decide that the cost is worthwhile as a form of advertising that helps raise awareness of your brand.
What about in the professional market? Do you think it was a worthwhile investment for ATI to build multi-GPU capability into the R3x0 and R4x0 lines? Do you think they're profiting from it?
 
Fodder said:
What about in the professional market? Do you think it was a worthwhile investment for ATI to build multi-GPU capability into the R3x0 and R4x0 lines? Do you think they're profiting from it?

I think the professional market probably needs to go beyond SLI and into the more esoteric world of multiple VPU cards like the stuff used at Evans & Sutherland, ie a dozen or more VPUs.

SLI is not really meant for the professional visualisation market - both Nvidia and ATI have other, more effective solutions for that market.
 
Fodder said:
What about in the professional market? Do you think it was a worthwhile investment for ATI to build multi-GPU capability into the R3x0 and R4x0 lines? Do you think they're profiting from it?

Personally I think its somewhat niche in the professional space.

As far as CAD/CAM users are concerned, vertex performance is more likely to be an issue, not pixel pushing power - and this is where the development of a unified shader architecture is more important than dual board use. The visualisation and simulation market does require lots of rendering power for high resolution displays and high degrees of FSAA, but this market is catered for by the likes of Quantum, SGI and E&S - all of these provide massively scalable systems and generally take care of the system infrastructure and software themselves. Digitial Content Creation may be one areas that can see larger benefits, but they are often offline anyway, so you are probably looking at accelerating the production process.
 
I don't see why anyone would feel that SLI is a "bad" idea. I don't have 400 bucks right now, but give me the option of buying a 200 dollar card right now and get another in the near future, then I might upgrade sooner than I would if I was waiting for prices to drop to lower levels.
 
ondaedg said:
I don't see why anyone would feel that SLI is a "bad" idea. I don't have 400 bucks right now, but give me the option of buying a 200 dollar card right now and get another in the near future, then I might upgrade sooner than I would if I was waiting for prices to drop to lower levels.
The problem I have with that line of reasoning is that by the prices drop to a lower level what else is going to be available then?

I don't think SLI is a bad idea at all, but I don't think it's going to be as feasible/easy as nVidia makes it out to be until dual-PCI mobos are a whole lot more common.
 
Karma Police said:
But the dual-PCIe mobo's are already on their way out in a few months from nVidia, Via, & Sis.

Yep, there is a strong dependency on the proliferation and relative cost of dual-PEG enabled platforms for the success of SLI in a broad market segment. But honestly, I only care about my experience and not whether every joe blow can afford or appreciate an SLI setup.

People gripe about the added cost of such platforms. But when I think about all the money I throw away anyway on other non-essentials like arctic cooling silencers, case fans, 500W modular sleeved PSU's 8) etc etc, then spending $30-$50 bucks extra on such a board is no big deal for me. Once you add in the cool factor the extra cost argument kinda goes out the window - it's just part of your hobby budget ;)
 
Karma Police said:
But the dual-PCIe mobo's are already on their way out in a few months from nVidia, Via, & Sis.

So in a few months. Its certianly nothing to talk about now. If in a few months the dual peg boards only carry a small premium over single peg boards that will still be paying more upfront for a system that later on you might want to put a second card into .

Then you have to factor in that its almost 6 months after the introduction of the 6000 series , which means the refreshes are almost upon us and its kinda dumb to buy one of the original sli boards . Then factor in your only going to have mabye a half a year or so for the board that you bought to drop before you have r500 cards and mabye nv50 cards which will have much better performance and better feature sets at the same prices.

I just don't see it very fesiable .



I also don't under stand what dave is saying. I wouldn't want to put an add in card inbetween two 6600gts , that would just kill cooling , i wouldn't want to put one directly under one either.

I don't even see maxx tech as a great idea (AS i've said before) but its at least better in the fact that its on the same card , its only going to take up 1 or 2 slots , you don't need a dual peg card .


Only real problem is that you have to spend the money all at once .
 
On that note I propose we stop talking about UE3.0, DX10, dual-core processors, xbox 2, EQ2 and everything that isn't going to be available next week :LOL: :LOL: ;)
 
Karma Police said:
But assuming ATi never releases SLI based cards, that means that 2Q we will see the R520 against 2 refreshed 6800's right?

Why would reviewers compare the R520 to 2 refreshed 6800's? The cost alone should put both solutions in totally different categories. The comparison should be between R520 and a single NV4x card.
 
trinibwoy said:
Karma Police said:
But assuming ATi never releases SLI based cards, that means that 2Q we will see the R520 against 2 refreshed 6800's right?

Why would reviewers compare the R520 to 2 refreshed 6800's? The cost alone should put both solutions in totally different categories. The comparison should be between R520 and a single NV4x card.

exactly .


What i can see them benchmarking would be

2 6800non ultras as they will most likely be had for around the price of a single 520 or nv4x card .

The only problem with that is not only does it show ati in a bad light but i would show nvidia in a bad light.

You would never compare a 6800ultra sli vs a x800xt pe. Because one is double the price of the other .
 
jvd said:
Then factor in your only going to have mabye a half a year or so for the board that you bought to drop before you have r500 cards and mabye nv50 cards which will have much better performance and better feature sets at the same prices.

Why would you factor in the R500 and NV50 cards when the dual PCI-E boards come out? Those cards would obviously not be out yet at that time, so it's obviously just going to be a tech demo right? In fact, why discuss R500 and NV50 now - they're certainly nothing to talk about now nor when the dual PCI-E boards are available.
 
dksuiko said:
jvd said:
Then factor in your only going to have mabye a half a year or so for the board that you bought to drop before you have r500 cards and mabye nv50 cards which will have much better performance and better feature sets at the same prices.

Why would you factor in the R500 and NV50 cards when the dual PCI-E boards come out? Those cards would obviously not be out yet at that time, so it's obviously just going to be a tech demo right? In fact, why discuss R500 and NV50 now - they're certainly nothing to talk about now nor when the dual PCI-E boards are available.

Why are we talking about sli right now , i can't buy an sli system so we should lock the thread.

The r520 cards are not all that far away , some are saying perhaps first half of next year , which is only about 8 or 9 months away ,if dual boards peg boards don't come out till around x mass , its just going to make the sli window on those cards smaller.
 
trinibwoy said:
He said in a few months digi :rolleyes:
nVidia has a nasty habit of paper launching products a whole lot longer ahead of the actual product roll-out, I tend to not trust their timeframes too much. ;)
 
jvd said:
Why are we talking about sli right now , i can't buy an sli system so we should lock the thread.

The r520 cards are not all that far away , some are saying perhaps first half of next year , which is only about 8 or 9 months away ,if dual boards peg boards don't come out till around x mass , its just going to make the sli window on those cards smaller.

I expect a similar request from you to lock all those EQ2 threads that you started since I can't buy EQ2 now either.

Also, are you saying that the R520 will beat a dual 6800 GT/Ultra setup? I wouldn't want to believe that your fervent objections to SLI are founded in the fact that Nvidia provides it and ATI does not ;)
 
Back
Top