Quote from the Geektimes article (does not necessarily reflect the opinion of the poster. )
Quote :
We may have to wait until December to see a Radeon 9900 with 256 MB of RAM. nVidia still claims that it will maintain a frenetic pace of GPU development, essentially introducing a new GPU core every year. If nVidia doesn't follow ATI's lead and somehow finds a way to increase its R&D budget 20% every year, ATI could find itself in a desperate situation. nVidia's NV35 should be faster than ATI's Radeon 9800, and nVidia may not provide its competitor with another opportunity to dominate the high-end GPU space. More importantly, nVidia will be able to migrate the high-end parts down to mainstream and budget levels more quickly than ATI. ATI could therefore lose marketshare to its archrival in every market segment. nVidia has always said that its plan was to own every pixel on the planet, and that dream may now be one step closer to fruition.
All right. Good idea, or bad idea for ATI?
<-----Insert witty sig line here.
<P ID="edit"><FONT SIZE=-1><EM>Edited by Twitch on 04/10/03 06:38 PM.</EM></FONT></P>
Im not saying whether its good or bad for Ati, but its great for us. I HATE upgrading then being left in the dust a few months later.... Their sales would increase heavily for one card if people knew it would be the shiznit for a long time come. Why not do that and make people happier, and give us something to look forward too while we enjoy having the best.
YEah slowing progres makes current owners feel good about their investment, but doesn't really benifit the consumer, and these are not investments, they are tools, despite some people investing alot of their hard earned cash.
I will believe the reports as the dates approach and as actual competing products reach the market, heck I'm still waiting for the R9600, and THAT release date has actually been pushed UP not back.
Anywhoo, we shall see. Hopefully it's just rumours. I just hope that OTHER card maker add to the segment. The MAtrox 'SunDog' or Sis 'Xanadu' or something (will we ever see the P10?). I still think DREAMCATCHEr would be the best name for a Graphics card line (of the name and not number variety).
- You need a licence to buy a gun, but they'll sell anyone a stamp <i>(or internet account)</i> ! <font color=red>RED</font color=red> <font color=green>GREEN</font color=green> GA to SK
The way I see it, games are so far behind the hardware that we probably won't even notice. The R9700 has been out for a long time now yet when is a dx9 game coming out?
24 months is 2 years, which is excessively slow. I won't believe 24, I will believe much less.
However the advantage I see here, is that competition in the sector of pricing will be stronger. Here's why: When companies release new products, sometimes they will add them to their existing list, trickle them and cause price drops on the other products. Lately, they have been refreshing the product lines, therefore replacing the cards instead of dropping them in price. That isn't necessarily good.
With a slowdown in card releases, nVidia may find itself in a situation where it thinks ATi has stopped competing, and hence will go monopolize the market, thinking they now own it, so they will either release a lot of products, or raise prices. ATi could from then just play the price drop game. So, I see this as a good thing to slow the cycles, as we can participate in a real price way where no new product is there, just the existing ones becoming a dog's treat.
--
This post is brought to you by Eden, on a Via Eden, in the garden of Eden.
Since this new 24 months product cycle does not seem to be a rumour but a fact according to businessweek.com, this will probably means the end of ATI success story as a market leader... if nvidia manages to keep up with their 6 months revision product cycle and 12 months new architecture product cycle!
Of course ATI can push their R3x0 core yet another bit beyond the R9800 Pro but can it keep up with the nv35 or maybe even the nv40 which could be released months before the R400 in Q1 2004, if nvidia continues its current product cycle track.
Before any unit01 or whoever again accuses me of NV PR BS or whatever, let me say that I don't like this 24 months product cycle of ATI and that this won't be good for any customer. But this again shows that even after many months a big success for ATI as also for AMD, those companies have not gained enough profit or size to be really competitive over a longer period...
nvidia can survive the failure of the nv30 without any major problem, but for a company like ATI (not saying that is little joe's company), it's a different question how good they will survive a year of performance crown absence... well, they did it before the R300 so probably it could work again this time.
To really see how dramatic this 24 months decision can be for ATI, we still have to wait for the performance of the nv35 to be sure...
History repeats itself and maybe this is really a second AMD story as with many AMD fans who no longer bought AMD CPUs after months because of the lack of any new and really exciting products. Hopefully not
It could be that ATI just is not impressed with their revenues at the high end of the graphics card market. I could see a scenario where Eden's suggestion might function--where ATI has a high-end product for 3-6 months, then gradually lowers the price on that product. For example, 9800 PRO is the hot ticket at the high end right now. Now, for the sake of argument, let's speculate that NV35 will be superior across-the-board when released.
Instead of rushing to market with a new release, ATI's strategy will be to lower the price on the 9800. So, if the NV35 is 20% faster overall, and costs $400, ATI will lower the price of 9800 PRO to, say, $299. In other words, ATI will concede the high-end to nVidia the majority of time in exchange for lowering their own R&D costs and giving themselves the ability to better support each release. Perhaps, in this way, it will be ATI who acquires the reputation for performance AND stability over the long-haul.
Now, I can also see a scenario where this can burn ATI. If NV35 is more than just slightly faster than 9800 PRO (say 40%-50% in some tests) and nVidia releases a mainstream part that is competitive with ATI's high-end, things could get very messy for ATI. In that scenario, nVidia would be benefitting from owning the high-gross-profit market, allowing them to continue putting money into research and development. Then, it would simply be a matter of continuing to compete at the low-end, and ATI could find itself out-of-the-loop in the vast majority of systems.
My biggest fear for ATI is that they have overestimated the impact they have had on the market in the last year. Yes, they have definitely made a lot of waves, but I would wager that at this moment, if you asked 100 ordinary consumers (read: not enthusiasts like us) who makes the best graphics cards, at least two-thirds of them would say nVidia. In this respect, nVidia is somewhat like Intel--great name-recognition.
That's why I find it so difficult to believe ATI would really handcuff themselves this way. I think the 24-month product cycle is theoretical, and will rapidly be round-filed if NV35 proves to be a superior product to 9800 PRO. In that case, I don't think ATI will hesitate to expedite R400 to the shelves. However, if NV35 is disappointing, there really will be no reason in ATI's mind for them to rush a new product to market, and they can proceed according to plan with the H2/04 release.
Poster: Twitch
Subject: Re: No R400 until Summer 2004.
It could be that ATI just is not impressed with their revenues at the high end of the graphics card market. I could see a scenario where Eden's suggestion might function--where ATI has a high-end product for 3-6 months, then gradually lowers the price on that product. For example, 9800 PRO is the hot ticket at the high end right now. Now, for the sake of argument, let's speculate that NV35 will be superior across-the-board when released.
The problem is that only the company which has the top performing card (whatever the price) is selling most low- and mid-end cards because customers don't want to buy cards from the second fastest manufacturer. For the time being everyone knows that ATI has the fastest card so most people will go for a Radeon 9500 or Radeon 9600 because the company is currently the leader.
I believe that would be the main problem with ATI that it can't offer the fastest card and people will again turn away from them and buy mid-end nvidia cards because they are "better". You don't make any money in the high-end sector, you just need it to establish your brandname.
ATI and nvidia are well aware of what this leadership position brings them in terms of marketing even if no one can buy their fastest card --> GF FX 5800 Ultra. Without the disappointing FX, nvidia would have certainly more problems at the time being then they now currently have, even if the GF FX was not so well received.
That's why I'm skeptical that ATI will actually stick with this plan. From what I've read, NV35 will be significantly faster and better than NV30. Since NV30 is competitive with 9700 PRO, and 9800 PRO is only slightly faster than 9700 PRO, I tend to think NV35 will be better overall than 9800 PRO. I've always thought it was a mistake for nVidia to rely on DDR2 memory rather than moving to a 256-bit memory bus. They appear to have remedied that problem (and several others) with NV35. The NV30 appears to have a lot of potential, but is throttled by numerous bottlenecks and driver issues. *IF* nVidia can fix the majority of those problems, NV35 could be a great product.
And I think *IF* NV35 is a great product, ATI will scrap the 24-month product cycle and expedite R400. I just don't know if a 256 meg DDR2 9900 PRO would be that much better than a 9800 PRO. I mean, the 9800 PRO isn't exactly hurting for memory bandwidth, from what I've read.
If NV35 is a superior product, it would be quite unwise for ATI to wait until summer 2004 to release R400, IMHO.
That all being said, we won't really be able to make any judgements until NV35 is released.
On the positive side, this might mean that Ati would pour more resources into chipset refreshes, like they did with the 9700 to 9800. Adding small features and increasing clockspeed every 6 months.
NV35 will kick ass I'm sure, but I'm interested in seeing what the R300 core can do with it's 256 bit memory bus paired up with DDR2... even if NV35 beats an R300 monster like that, I'll bet it won't be by a huge margin.
Looks like time will tell though. But I have to say it doesn't bode well. When all is said and done, Nvidia has wayy wayy more $$$ to throw at R&D, and they're determined to be the top dog again.
Ati might be right to settle for second place if it means a boost in profitability...
NV35 will kick ass I'm sure, but I'm interested in seeing what the R300 core can do with it's 256 bit memory bus paired up with DDR2... even if NV35 beats an R300 monster like that, I'll bet it won't be by a huge margin.
There won't be a Radeon with 256-bit DDR II for a very long time, probably not earlier than 2004. DDR II is still too expensive and 256-bit DDR II is even more expensive. This card could not be reasonable priced...
ATI has probably switched to a 24-month product cycle because nVidia has been forced to do so too. The last major card before the GeForce FX was the GeForce 3 about 2 years ago. ATI is merely slowing down to compete with nVidia. ATI doesn't need to spend extra on R&D when they can just rehash their existing products. Technically, ATI is on a 6-month product release cycle where they release a new card or set of cards every 6 months with a slightly revised core. ATI promised this with the release of the R9700Pro and so far, they're living up to their promise.
Intelligence is not merely the wealth of knowledge but the sum of perception, wisdom, and knowledge.
)
GA to SK