So, ATI is doomed..?

digitalwanderer said:
We'll have to agree to disagree then Uttar, I think good deals should be mutually beneficial to both parties for true company growth.

nVidia might have profited well from M$, but they ain't gonna be doing business with them again any time soon....ATi's deal seemed smarter to me and thus sweeter in the long run.
I think companies will cooperate with each other as long as it's mutually benficial (ie profitable) and rarely will silly things like "principals" come into any business decisions. The issues MS and NV had probably played some part in MS seeking out ATI, but I'm sure that if the right opportunity comes along both companies will find a way to work together again.
 
Yeah, as I understand it (pretty vague, admittedly), M$ and NVidia worked together on HLSL, with NVidia contributing quite a lot.

Although this then leads into a discussion about Cg "competing" with HLSL and it all gets very messy. Hmm...

Jawed
 
digitalwanderer said:
nVidia might have profited well from M$, but they ain't gonna be doing business with them again any time soon....ATi's deal seemed smarter to me and thus sweeter in the long run.
AFAIK, it's a bit more complicated than that. Their original contract didn't have anything that made NVIDIA profit more than MS, and NVIDIA's margins on the XBox chips were low initially. That's not the problem. The problem is Microsoft had assumed there would be price renegociations as components get cheaper, as tends to be standard in this industry. NVIDIA refused to renegociate.

Microsoft then sued NVIDIA over this, and NVIDIA basically won the lawsuit. Officially, they found a mutual agreement, but Microsoft gained little from it, and NVIDIA apparently didn't have to concede much of anything either.

ATI's contract, on the other hand, is made with "no renegociation" in mind: the chip costs are Microsoft's business, and they will simply pay a flat fee to ATI for each XBox sold (or shipped, or manufactured - I don't know for sure about that). ATI most likely has equivalent or slightly higher per-console profit than NVIDIA had at the beggining of XBox1 production, but they cannot dream of the per-console profit NVIDIA was getting in the console's last days.

Ironically enough, the PS3 contract NVIDIA has engaged to is roughly similar to the XBox360's. IMO, it could be argued that the renegociation had to happen at a time where NVIDIA was in an extremely annoying position with regards to NV3x. They had to squeeze every bit of profit out of "successful" products they had in the last year. The same could imo possibly be said of the slower-than-expected GF4 price drops post-R3xx avaibility.

Had NVIDIA given a flying fuck about their relationship with Microsoft's console division in that timeframe, their contract would have had no less problem than ATI's current one. The only real advantage of ATI's contract that I'm aware of is that even if they wanted to, they couldn't really use this to damage their "relationship" with Microsoft.

I am actually unsure of who handles respins in the licensing model both MS and Sony use for their GPUs; obviously, the IHVs have no reason to waste any resources on them, and as such I would assume that to be handled by those companies' engineers instead. I'd love to have some further information on that, however.

Finally, let me insist that NVIDIA's "problems" with MS, as Ratchet explains, aren't really "fundamental". Furthermore, many factors contributed to them, among which some related to NV3x. I'm not going to get into the little I know about that here, but hopefully all of the above will clarify my earlier statement(s).


Uttar
 
Bottom line, though, is that for whatever reasons the Microsoft-nVidia relationship did not work or last. My own opinion about that is that nVidia wanted to pull all the strings, chart the course, and basically usurp much of Microsoft's volition in the entire xbox concept as it progressed. nVidia wanted the nV3x generation of nVidia concepts and chips to push the xBox forward in terms of not only graphics, but core-logic support and sound support--basically, the whole concept was something nVidia envisioned itself taking over (imo.) We saw the strain evidenced back in '02 and into '03 when nVidia started publicly arguing with Microsoft's decisions relative to DX9, for instance. Further division evidenced itself with the disputes over how nVidia felt Microsoft should pay nVidia for its role in the xBox. Etc. Basically, I think nV jumped the gun prematurely in so many ways by assessing its position in a great many things to be much more secure than it actually was. Including the xBox.

What happened is that nV basically wanted to milk its existing DX-8-ish era technologies for far longer than Microsoft had any interest in doing, and there I think was the "rub"--the straw that broke the camel's back in the relationship between the two companies. nVidia decided it was in a position to dictate to Microsoft on a number of fundamental things and Microsoft disagreed proactively. And there you have it...;)
 
Thing is, nVidia was basically correct.

DX9 was essentially irrelevant to games until NV40, R4xx and Half-Life 2 came along.

Up until then DirectX 8 was what developers targetted, in no small part due to the fact that the original xbox was also DX8 and the huge install base of Dx8 generation graphics cards already out there.
We will see the same thing with the new consoles IMO, DirectX next (or whatever you want to call it) will arrive alongside Vista, however it will be largely irrelevant to gamers featurewise for at least 18 months due to the install base of DirectX9.0C video cards and both major consoles having DirectX9.0C level graphics capabilities.
 
radar1200gs said:
DX9 was essentially irrelevant to games until NV40, R4xx and Half-Life 2 came along.
Had there been two quality DX9 architectures on the market, might it not have arrived in force sooner?
 
No, I don't think so. The simple fact of the matter is DirectX 9 was released almost 2 years before it was truly required.

DirectX 8 was still a young and powerful API that had barely been touched by games develeopers and was just starting to come into its own.

It is only now being gradually replaced by DX9, but it will be a long time yet before it is totally dead (support won't drop off until the new consoles are at least one year old), nad even then, it will remain the base level for graphics - developers will write 2 versions of their code at most - one DX8 level for legacy support and one DX9.0C level for currentl hardware and consoles with nothing in between.
 
radar1200gs said:
Thing is, nVidia was basically correct.

DX9 was essentially irrelevant to games until NV40, R4xx and Half-Life 2 came along.

That's no help if you are a the current owner of a FX5900U and you play HL2 or farcry. Then watch your 5900U beaten by a 9800XT in doom3 benches.
Gamers who bought a 9800 back then or even a 9700 are better off.
 
I never mentioned anything about underlying hardware support, did I?

It's good to keep the hardware evolving, even if the API doesn't keep pace, it allows developers to become familiar with the future API (so when the new API does launch people don't have to wait 18 months to see games appear that exploit the API fully) and consumers benefit with things like AA/AF and speed.

By the way, I own a 5900XT along with my 6800GT, and the 5900XT manages to play HL2 (in DX9 mode thanks to a user made patch), farcry and Doom3 perfectly well, thanks very much. Not as quickly as the 6800GT granted, but then the 6800GT is nearly 1.5 years younger than the 5900XT...

EDIT:
Link to the HL2 patch
http://translate.google.com/transla...n&ie=ISO-8859-1&safe=off&prev=/language_tools
 
Last edited by a moderator:
I agree with Fodder. It seems to me that if nV's FX series were as capable as ATI's in terms of DX9 performance, we would've seen more DX9 effects sooner, especially as both IHVs pushed "DX9" cards into the midrange (9500, 5600) fairly quickly, and nV, into the mainstream (5200).

DX8 may well have been young and powerful, but DX9 was as powerful and--thanks to HLSL/Cg--easier to code for, no? And devs probably would've tossed in the early DX9 effects they created while becoming familiar with the API, similar to how we saw HDR implemented before it was usable by the majority of the market (consoles included). At the very least, it would drum up some extra press for their game.

In my amateur opinion, that is.
 
Maybe. Maybe not Pete. My reasoning for this thought dwells on the LCD (Least common Denominator )) factor that tends to influence devs in their design. Even our Defacto DirectX 9.0 game really offer limited/marginal improvements over their baseline DX 8.0 features. We may have seen a few more DirectX 9.0 effects in current titles. But I dont see a that it would be any different from our current DX 9.0 titles where it offers little over the DX 8.0 implementation. ((little to the average user. Discriminating users obviously view things differently))
 
digitalwanderer said:
Actually I think they Walt was referring to the contracts made with M$ by nVidia & ATi, ATi has a much sweeter/smarter deal set up with M$ than nVidia so hopefully it won't end in a crash-n-burn relationship like M$ & nVidia did. ;)

afaik Nvidia ended getting paid what they were deserved per the contract and not some lower figure Microsoft was proposing because they missed sales by a long shot.

So the company who ended up with the not so sweet deal out of it was Micrsoft because their sales projections were nowhere near where they needed to be.
 
I would also not advise to think a company is making profit from selling a lot of products or won a contract. Two cases:

Best Buy or Circuit City selling computers. Margins are so thin, that they usually lose money from selling computers that are in the ads.

Any type of contractor taking a bid on a product. If they cannot complete the project under the amount the contracting company bid for a product, then they will lose money completing the project.

Without knowing the specifics, we cannot say whether taking the XBox360 deal was a good "financial deal" for ATI. The deal definately gives ATI good exposure, and sets up future deals, but it may be a low profit deal. In fact, the deal may have a been a bad idea in the short term. We just do not know how much ATI had to cut to make Microsoft happy (granted I dont' think Nvidia seemed to compete for it). However, ATI believes the deal was good or else they wouldn't have taken the contract.
 
Back
Top