To start, your GameCube analogy is incorrect. GameCube doesn't play all older Nintendo games.
Exactly. I wrote: Imagine the GameCube would be out now and that it could play Nintendo64 games. Imagine.
If I had a GameCube right now, there'd be pretty much zero that I could do with it.
What I was trying to say was: IF Nintendo would release the Gamecube NOW and IF it would be able to play older games, it would be the same situation as with the GeForce 3 - except that everyone knows that:
- Nintendo wouldn't do something like this
- Nobody would buy it
The console industry works differently. The GF3 is a component. A video card. An important part of, but not a whole, platform. This is a core reason why the industry works in a different way than consoles.
The weird thing is: The GeForce 3 *will* be bought.
*My* point exactly. If I was putting a cutting-edge computer together right now, I could buy a GeForce2 MX or GTS, and get a GeForce 3 later when they're cheaper. That means $100-$150ish right now, and some amount later, depending on how long I want to wait for the GF3. There's little competition with nVidia for the high-end gaming market, and I don't think the GF3's price will drop very fast very quickly. Look at the GF2 Ultra - it's their current top-end card, and in its several months, it's gone from $500ish to just under $400 - not a big change.
Now, in my hypothetical new computer - I could buy a GF3 right now. Which one is cheaper in the long run? Depends. But personally, I'd probably just go straight with the GF3 - I'm pretty sure that they'll be seen in the (probably high) $400s by release time. In the end, I likely could save money this way, and in the meantime, I'd have a faster card than a GF2 MX/GTS.
On the other hand, the GeForce 3 works with older PC games. If I had a GeForce 3 right now, I could use it. Not to its full potential, of course, but it'd work, and it'd work roughly as well as a GF2 Ultra.
Also, you may have been misreading recent benchmarks. In fact, did you even see them? It's hard to believe that you did, especially after you said this: "Why ? What's the point in getting from enough frames to more than enough frames per second in Quake 3 ?" Um, it's no faster than an Ultra. It's basically just more features.
Well... The benchmarks we've seen of GeForce 3 boards so far don't have to be correct - nVidia questioned them. All those were pre-production sample-boards.
This seems to happen with just about all video cards. Before release, someone sneaks out some benchmarks, and the manufacturer writes off those benchmarks because they don't show a revolutionary speed increase. Unless I have my history wrong, it just about always wound up that the card really *was* that fast, the drivers didn't make a very noticeable speed increase between then and release, and people forgot about it anyway. I seriously doubt that the GF3 will show any significant speed differences by release.
But even if it is not faster than an Ultra, it proves my point that there's no reason to get one for quite some time.
Depends on whether 1) you already have the computer to push that kind of video card and 2) whether you're upgrading to a GF3, or buying a new one for a new machine. I think the GF3 is the right choice for a new computer, but I do agree that it's a waste as an incremental upgrade from a GeForce 2-class card in a mid-to-high-range computer. In fact, it might be a waste for quite some time - maybe even until next year. I doubt the prices are going to drop too much, since nVidia has little to worry about from ATI or *chuckle* 3Dfx.
Thanks for your feedback.
I'd imagine you'll be getting some sort of hateful flames from some people ... : )
And unlike consoles, it's a standard thing for video cards to be released, the early adopters pay too much for them, and then games come out that use the new features on the card months later. That's the norm. If you're complaining about that process, then don't just complain about the GeForce3 - this happens to any company who has been innovating hardware over the years.
Of course, buying a GF3 right now is a waste of money. That's a given - buying the very newest stuff is always like that in the PC industry, because they offer only incrementally better performance for sometimes double the price of the previous product.
At least in the case of the GF3, instead of just more and more frames per second, it has a lot of new features that will make future games look much better (at roughly the same frame rates as a GF2 Ultra). Which we've needed for quite a while - not a lot of astounding new features have been seen in 3D cards in years (well, in my opinion). I'm actually glad that nVidia decided to jump off the frames per second bandwagon, and make a card that improves visual quality instead of frame rate.
As seen from early benchmarks the GF3 is not that blazingly fast as focused on in the article.
Where did I focus on that ?
The impression I got from the article as a whole (let alone the title) was that you focused the negativism on speed. (even though it was not the only issue brought up)
Really, The whole point with GF3 seems to be primarily focused on bringing new features to 3d graphics. These include 3d textures and in perticular, vertex programming.
In fact, with todays games the GF3 do not seem faster than the GF2U.
So what is the rant about? I really don't get it.
The "rant" is about the price, about the fact that nVidia's products seem to get more and more expensive and that some people actually buy the boards the moment they hit the shelves despite the fact that the features and the additional speed (if any) will not be used for quite some time.
If NVIDIA doesn't pioneer new 3d features who will?
The article does not question nVidia's position as the market leader and their extremely good work developing new chips and bringing new features to the market.
It's about the question whether it's necessary to shell out that much money just now.
Thanks for your feedback !
I do agree with you on the price issue... but maybe the title should have been Too Expensive For Us? :)
Still, it's probably good that someone discourages crazy people from dishing out their savings for 10 fps :)
Some thoughts on the GF3 and its high cost:
Doesn't piss me off it is so expensive. It just helps me to save that cash for things that I REALLY need right now. If nVidia waited to sell the GF3 until there were games to play on it, they wouldn't have any new product to sell for over a year. They could not keep putting out iterations of the GF2 technology, so a leap like the GF3 requires this sort of pricing strategy.
In the PC world, you can't wait for developer adoption of a card before you sell it. nVidia needs to recover costs for its R&D, which has been extensive. Nintendo has to ship out developer kits for its gaming platforms months before they put a console on the market, two reasons being that: nothing they build is backward compatible, and console revenue models are driven around game license fees, not console sales. In some regards, Sony almost did what you suggest nVidia is doing by releasing their console with few games designed for it that can be justified for its high cost ($300), but it is backward compatible with its older PS games and has DVD functionality which is a big reason why it sold. But no, you cannot make a parallel between console gaming and nVidia, since nVidia does not get license fees for the sale of each computer game that uses their card. They get general developer fees for those developing games for use on their cards, but then so do ATI, etc. Their cash recovery model is much different. And of course, consoles are priced generally to lose money for their manufacturers since the real money comes from license fees, and in the case of Nintendo, cartridge sales. You sell 1 million consoles at a loss, then earn up to 50% per game unit sold that use that unit. nVidia doesn't have this option.
So who will buy these cards? Initial demand will come from the gaming development community, who can write these cards off as a business expense and not feel the financial pain. There are always early adopters with cash in hand, but mostly professionals will get this card over the next few months.
Research, development, and production costs are higher for the GF3 than previous development efforts. Since the GF3 chipset is expected to have a healthy life cycle, they will skim the market with the initial release, then lower the price to penetrate the market further. This is why Creative and Abit are staying out right now, since they know they can't make much money on this given the price point. nVidia would not intentional screw itself with high price unless it knows exactly what it is doing. Starting out with a lower price means losing money. How can you piss off a market with a high price for a card that offers little current benefit? nVidia and the various vendors know what they are doing, and the market is fickle. People pissed off by the price now will still buy the card when it makes sense to do so.
Clearly the GF3 offers few advantages for current games over GF2 cards (except FSAA and faster 32bit color at high resolutions), so nVidia must not be banking on short term gains. nVidia's GF3 investment will pay off 12-18 mos. from now when the GF3-specific games start hitting the market, at which point the price will be more affordable and they'll start flying off the shelves when Doom 3, etc. see the light of day.
Finally, given that the GF3 technologies might result in future games that only work on the GF3 seems to indicate that nVidia has established its own playground with high barriers to entry by its competitors. When you have your own place to play, and everyone wants to be there, you can charge whatever you want. nVidia is just being smart. So, no, I don't care it costs so much. It just shows me that the GF3 is here to stay for quite awhile, and when the games are out that will rock on the card, so will be my wallet buying a more affordable GF3.
sorry, but what if your understanding of "enough frames"?? A game is only smooth when the minimal framerate is equal the frequence of the monitor, so at least 60 frames per second!
In Quake3 there are two maps that aren't completely fluid with an overclocked GeForce2 Ultra, ColinMcRaeRally2 doesn't run at 60fps, Severance neither and so on and so on... Thre you have the need for a more powerful chip!!!
Besides, there are other advantages like better anisotropic filtering and antialiasing...
MfG [German abbreviation for "Greetings", Ed.]
- Marc :)
It's Tokyo, not Tokio.
Some folks buy Ferraris, some Hondas. Both get you from A to B.
Don't be mad.
For your info the geforce 3 can be pre-ordered at www.computers4sure.com for $490.00 (a lot less than $550.00 you wrote in your article) thats about the same as the ULTRA a few months ago. (by the way its the Gladiac 920 full retail). Besides Nvidia isnt stupid they'll lower the prices soon.
You couldn't have said it better! (well maybe...) But it's true, why by the latest card from nvidia, I still you my geforce one DDR and it kicks ass. I don't need a new card till mabye the geforce 5 :)
I couldnt agree more with you on this topic. I purchased an Nvidia Geforce 2 GTS 32mb DDR card over a year ago and game manufacturers are finally encorporating its transform and lighting features in games. This past year and the cards extra cost a year ago have basically been worthless.
I agree that for Q3 a GF3 is more than needed. However, let's point out a few things.
Q3 is not the only 3D game out there.
UT? Runs like 2 legged dog on my Geforce DDR. Not Nvidia's fault, but something that more speed may indeed help. Not that UT is worth $500 for a GF3 by any means. And no I won't be getting one until they are $300 or less myself, still..
Baldur's Gate 2? Turn on ALL the effects to max under a resolution of 1024x768 and get in a big battle. Ouch.
That's right, we can use the new power even in SOME of today's games, but what about the others? Obviously its no use for starcraft, not really useful in Diablo II, Red Alert 2, and probably not in Age of Empires II either. So......
Any 3D game will take advantage of two features of the GF3.
- Improved memory management. Yeah, no such thing as a graphic program that doesn't use the video memory, so......
- New FSAA sampeling, assuming you like blurry screens, I don't, but..
Though #2 may not be a GF3 only feature, it is possible the FSAA method can be implemented on the older GF cards, I suppose.
As I said, it isn't worth $500 to me, but if it were $300 or less, yes I would buy it, and yes, it would be useful even in today's games. Just a point to consider.