R580+

The 7900GTX is averaging $450+, the X1900XT is averaging < $400. (Sourced from pricewatch.com)



If you think marketing = TV commercial when it comes to 3D hardware I must say you're not really in a position to comment on it. Nvidia is everywhere in the gaming community - magazines, games, competitions, trade shows. They do everything bigger and better. It's not about ad banners - it's about presence in the community(not a reference to shilling :smile: ) and selling the advantages of your products and technology. Have you visited SLIZone or Nzone ? Have you seen the ATi counterpart? It's a joke.


Exactly. And let's not forget TWIMTBP (though I suspect that's partially what you meant when you mentioned games.) That, and ingame bilboards and logos spattered all over the place.
 
Huh? Nvidia is basically ATi's equal or superior at every price range.

There's a little image qaulity issue but most people dont care very much about that, since it's not something a layperson would even notice.

7900GTX is pretty comparable to a X1900/50 XTX. Same in SLI configs.

In most catagories NV arguably has the lead. 7900GT beats a X1900GT most of the time, especially when vendor overclocked as most NV cards sold are.

And to me where ATI really messed up, is they have nothing even CLOSE to the 7600 GT at $130 now. I mean they're not even trying to compete at that segment any more. Which is probably the most popular segment too boot. X1800 GTO was I guess the competition, but ATI's architecture missteps are such that the die size is huge for that part, including a 256 bit bus, so it cant compete on price, or even performance anyway, 7600GT is faster most of the time.

ATI has some nice new parts that put the perfomance pressure on NV in the 199-299 segments (again lacking the crucial $130 part though), but guess what, the 7950GT and GS will counter or exceed those in a couple weeks. All the while NV has the advantage of so much smaller die sizes the profit differences aren't even close.

And as well, NV cards use less power, which is probably easily as or more important to the average user than the small IQ issues that favor ATI. Electric bills aren't cheap these days, besides heat etc issues.

And I haven't touched on the 7950GX2, which like the 7600GT ATI doesn't have any real answer to at all.

Nope..not seeing any equal or superior ATI parts selling at lower ASP's here..by and large..

I will agree NV is generally more popular though, but if ATI was generally significantly faster, which they aren't, people would throw their NV preferences out the window. And with double size dies, there's not much excuse for ATI not being a hell of a lot faster than NV. But they simply aren't.

Too me marketing is usually one of the most overrated factors in business success. I've never seen an Nvidia commercial. The only IHV marketing I can think of at all right now is ATI's ever present Ruby chick. That's it. Cant think of any NV marketing at all. Dont even know what this marketing is..is it banner ads? Dont recall seeing any banner ads for ATI/NV either.

See this at tomshardware....
060823_gpu_analysis_chart.jpg
 
And to me where ATI really messed up, is they have nothing even CLOSE to the 7600 GT at $130 now. I mean they're not even trying to compete at that segment any more.

ATI has some nice new parts that put the perfomance pressure on NV in the 199-299 segments (again lacking the crucial $130 part though), but guess what, the 7950GT and GS will counter or exceed those in a couple weeks. All the while NV has the advantage of so much smaller die sizes the profit differences aren't even close.
Well, they are "trying", but like you just mentioned it, they didn´t have much of a choice WRT to their underlying architecture and their margin model. Like NV, ATi needs a product mix to achieve/meet their underlying terms of their own margin model, otherwise they would end up having drastically lower margins on one ASIC (SKU), which they invested good money in, which normally isn´t an option, but ATi had to face it anyway with R520, so their own calculations had been down the drain, just to give you an example why ATi didn´t have a lot of options then and now.

So, if you add everything up, it´s a calculation based on your whole product range, not just "one ASIC", which makes it pretty hard to compete, since you basically have invested in features the competition doesn´t have (currently), but you were late to the game. It´s like going on a tour for 6 months, coming back and find your wife with another guy, while you worked your ass off (i´m not talking about my own experience, fortunately) on your job and came back with nothing. You need to get sorted, but that takes time, time that´s very hard to get with competition always around, trying to bite you in the ass.

To sum it up, what NV currently does is just to solely rely on their time-advantage they had with G70, they try to keep this advantage by shrinking some ASICs again (therefore they can lower the prices even more, theoretically speaking) and wait for G80 to happen, it´s not much more than that.

Too me marketing is usually one of the most overrated factors in business success. I've never seen an Nvidia commercial. The only IHV marketing I can think of at all right now is ATI's ever present Ruby chick. That's it. Cant think of any NV marketing at all. Dont even know what this marketing is..is it banner ads? Dont recall seeing any banner ads for ATI/NV either.
It´s not overrated, it´s just misunderstood by a lot of people and companies. It´s a concept which only begins at the point of your own confidence, instead of just printing some charts and hand it to your target-audience. It doesn´t exactly work that way. Marketing is something that can - together with the companies' own philosophy - change the whole perception of everything, that is, if you do it right and your products are good enough to have certain advantages the competition doesn´t have.

Pre-D3D10 times were practically a mess, mud-wrestling if you will. Not implying that this will never come back, but according to the way D3D evolved, there´s no comparison. I´ve never seen anything like it in the PC-world, for the average gamer / customer it just was hella confusing. I know that because i worked for a very large HW-discounter in Germany, where i always had to communicate with people (not only single customers, but big companies) and tell them "why exactly" product A is better than product B. In the end, it always came down to price/performance and today it´s no different, even for people with high standards, they always care about price/performance and of course, hearsay, which is the marketing side of it.

Now, coming back to sole marketing:

Given that there´s always competition around and you want your own products to shine, where you really have an advantage, it´s game/gamer-related and not just some checklist-feature, you have to tell your audience or you shouldn´t have invested in it in the first place. Now, all HW-sites basically want to do their best in communicating those advantages to it´s target audience, but a lot of them don´t even understand them, they just put charts (which they were given from the IHV directly) on their page, but they don´t tell the typical layman how it´s exactly important and why you would need it, because that´s where good journalism begins.

Therefore, marketing is a lot more than only "advertising". It´s about your own confidence you have in your own products and you have to be so proud if it (think NV) that even if your products don´t have certain features, you still can tell the people "we´re the best, because we have way lower power consumption and lower prices". You have to communicate it right, even if typical perception is geared to other directions. It simply doesn´t matter if you are self-confident enough.
 
Last edited by a moderator:
Well, they are "trying", but like you just mentioned it, they didn´t have much of a choice WRT to their underlying architecture and their margin model. Like NV, ATi needs a product mix to achieve/meet their underlying terms of their own margin model, otherwise they would end up having drastically lower margins on one ASIC (SKU), which they invested good money in, which normally isn´t an option, but ATi had to face it anyway with R520, so their own calculations had been down the drain, just to give you an example why ATi didn´t have a lot of options then and now.

So, if you add everything up, it´s a calculation based on your whole product range, not just "one ASIC", which makes it pretty hard to compete, since you basically have invested in features the competition doesn´t have (currently), but you were late to the game. It´s like going on a tour for 6 months, coming back and find your wife with another guy, while you worked your ass off (i´m not talking about my own experience, fortunately) on your job and came back with nothing. You need to get sorted, but that takes time, time that´s very hard to get with competition always around, trying to bite you in the ass.
Well, then what is ATi doing in meantime?
For more than a year they surrendered the 100-200$ segment, with x1600 not having a chance vs 7300Gt/7600 . Why so late with 8TMU parts ? i don't want to touch the choosing of 3:1 alu/tmu ... ok, they made wrong choice, but what was the reasoning to have 4 and 16 TMU parts (ok, add 12 one using "bad" chips with 16) ?!
i don't think thats because of marketing. What has marketing to do woth choosing number of pipes?! Its top management, the people who take decisions who misstepped.
 
Well, then what is ATi doing in meantime? For more than a year they surrendered the 100-200$ segment, with x1600 not having a chance vs 7300Gt/7600 . Why so late with 8TMU parts ?
Like i mentioned, there definately are different reasons, but mainly because they chose a way which - the architecture side - had them already at a disadvantage, apart from their totally messed-up internal timetable. It think they haven´t really well-communicated what really was important back then (and what isn´t) and their philosophy was "too advanced" and they looked a little "too much into the future" and/or they just wanted to stear the industry in another direction which they thought will give them an advantage in the end. That´s the perception side of it, because you have to tell people what exactly is your advantage and it should be something that is worth either shelling out more money for and/or worth owning in the first place.

Now, speaking of the ASICs itself. No IHV can really justify putting out ASICs that lower their margins to an extent where it doesn´t make sense to compete in the first place, because it would totally mess up their margin model, at least that´s what i think, because as you already wrote, from the outside it looked like they didn´t do anything, apart from introducing SKUs that are based on defective, disabled or purposedly well-working R520/R580 cores in their lower SKUs, which is - frankly - the only thing they could do and thats also why you certainly hit the nail on the head here.

All of this typically needs to be done in an offset-manner, meaning that less complex ASICs also typically give you both higher yields, good margins and you can also sell a lot of them to earn money, that´s how NV does it and they really showed that it works. It also has the very positive side-effect that you can react to your competition very quickly in terms of new SKUs you want to place into the market without beginning from scratch. This is all a result of their underlying architecture, the scalability of it.

Since RV530 already is a fair bit above G73 in terms of die size and performance is worse, they either had the option to go with 90nm or 80nm and they chose the latter, because they can dedicate more die space for certain things that are performance-relevant while their margins don´t take another hit. However, as everyone should know by now, 80nm wasn´t really "ready" until recently, so we have another time-related factor here that obviously didn´t work out well for them.

i don't think thats because of marketing. What has marketing to do woth choosing number of pipes?! Its top management, the people who take decisions who misstepped.
I think you misunderstood what i meant or you didn´t read my whole 2nd paragraph, which was adressing this. What i´m saying, is that marketing is the "gap" between a typical customer and the company itself, which is critical in terms of communication. You need to close it and profit from it, rather than letting the gap speak for itself.

Management is a very complex topic altogether, so i´m not really gonna question one companies' internal decisions, because obviously no one really knows all of the hard facts. (apart from Dave Orton, etc.)
 
Last edited by a moderator:
i don't think thats because of marketing. What has marketing to do woth choosing number of pipes?! Its top management, the people who take decisions who misstepped.
Marketing's job is to contribute in choosing the number of pipes and other features. Left on their own engineers will build some really cool products that might not sell. It's marketing's job to talk to customers and steer the decision makers in the right direction.
 
The 7900GTX is averaging $450+, the X1900XT is averaging < $400. (Sourced from pricewatch.com)



If you think marketing = TV commercial when it comes to 3D hardware I must say you're not really in a position to comment on it. Nvidia is everywhere in the gaming community - magazines, games, competitions, trade shows. They do everything bigger and better. It's not about ad banners - it's about presence in the community(not a reference to shilling :smile: ) and selling the advantages of your products and technology. Have you visited SLIZone or Nzone ? Have you seen the ATi counterpart? It's a joke.

Just curious. What is the ATI counterpart? I was completely unaware they had a website comparable to nzone.

Chris
 
Back
Top