Sign in with
Sign up | Sign in
Your question

Is there a good scenario for nVidia?

Tags:
  • Graphics Cards
  • Nvidia
  • ATI
  • Graphics
Last response: in Graphics & Displays
Share
March 26, 2010 5:48:16 AM

disclaimer: I don't think I am 100% objective on this subject. I have a few AMD stocks and I do favor ATI over nVidia. On the other hand I try to be open minded and I don't consider myself as an expert so I know I may be missing something or just dead wrong.

So here is how I look at the whole Fermi situation: tonight we will finally get some real info about the Fermi cards and will know how good it is but I think that for nVidia it doesn't matter, they lost this round anyway from the business point of view. This is why:

worst scenario for nVidia: Fermi is not so good. Performance is the same as ATI 5xxx. it's hot enough for BBQ. It is overpriced.
The result:
1. Only a few fanboys buy Fermi
2. ATI does nothing, prices don't change.
3. ATI sales skyrocket with nice margins
4. ATI wins

best scenario for nVidia: Fermi is super great. Performance is 30%+ more then ATI 5xxx. TDP is not so high. Price is very competitive.
The result:
1. nVidia wins the very high end market
2. ATI is forced to cut prices down of it's high end cards and therefore all others
3. nVidia has nothing to offer for the gamers other then very expensive cards so ATI wins the rest of the market with low margins
4. The consumers win

So nVidia can't win. Either Fermi is good enough to give us cheap ATI cards or we will have to pay full price for them in any scenario nVidia don't have anything competitive for tha majority of the market.

Do I miss anything or is it really the time to sell those nVidia stocks?

More about : good scenario nvidia

a c 217 U Graphics card
a c 83 Î Nvidia
March 26, 2010 5:55:32 AM

Fermi isn't going to be used on the lower end cards. They'll be refreshing their existing cards for that, and they'll perform fine, but lack DX11 support.

Anyways, the concept of Fermi's shaders being able to transform into tessellators or used for parallel processing is pretty cool. This allows the cards to be configured for the best needs of the software without having some hardware idle. If this doesn't work out great this time around, it still might pave the way for a future card that may blow us away.
a c 376 U Graphics card
a b Î Nvidia
March 26, 2010 5:59:05 AM

This is the worst case scenario for Nvidia;
http://www.semiaccurate.com/2010/02/17/nvidias-fermigtx...
The best case is probably what you listed above plus a 28nm refresh that increases yield and performance, decreases manufacturing costs and allows the architecture to be downsized into viable low/mid range cards.
Related resources
March 26, 2010 6:06:51 AM

bystander said:
Fermi isn't going to be used on the lower end cards. They'll be refreshing their existing cards for that, and they'll perform fine, but lack DX11 support.

Anyways, the concept of Fermi's shaders being able to transform into tessellators or used for parallel processing is pretty cool. This allows the cards to be configured for the best needs of the software without having some hardware idle. If this doesn't work out great this time around, it still might pave the way for a future card that may blow us away.


My point exactly. Fermi might be wonderful but ATIs strategy, to aim for the mid range and create from it high and low end derivatives allowed them to refresh all cards in a few month. nVidias monolithic approach gives them for the near future just very high end cards. Any DX10 card of nVidia will be less and less competitive as more DX11 games pop up .
a c 180 U Graphics card
a b Î Nvidia
March 26, 2010 5:45:09 PM

If you're deciding on your stocks based on such a simplistic analysis of NVidia as a whole you clearly don't understand how to analyze stocks.

NVidia is a lot more than just how its latest DX11 offerings do in the short term.

NVidia has Tegra and Tegra 2 for netbooks, media players and even the Nintendo DS I think. They'll probably have an investment in one or more of the future consoles (2012?).

NVidia has investments in computers for scientific firms. You should read about the NVidia relationship in developing the movie Avatar.

To understand about stocks for NVidia you need to do a LOT of research on Intel, AMD/ATI, NVidia, Google and Apple.

You need to understand most of the technology, the history of technology development, future trends etc. It is complicated.

Apple, for example wasn't doing very well then they came along with the iPod which not only saved the company but paved the way to the iPhone (Apps Store) etc.

A good company is diversified without being cash and research poor. NVidia is poised quite nicely but deciding when to buy and sell Stocks is very risky in any company and not a good idea if you don't have a firm grasp on the entire Technology field that NVidia is part of.
a c 376 U Graphics card
a b Î Nvidia
March 26, 2010 7:01:57 PM

I'm not so sure about that. If the Fermi cards are a big and highly publicized failure the stock prices WILL take a hit, possibly a large one, even if it doesn't make sense from a broader perspective.
a c 107 U Graphics card
a b Î Nvidia
March 26, 2010 7:18:58 PM

While it may be a bit too late for the GTX 480 and 470 cards, it depends on nVidia's management if they will go up from here, or keep tumbling down. The first derivative of Fermi is rumored to have 256 shader cores, 64 TMUs, 32 ROPs and it will utilize a 256-bit

http://www.fudzilla.com/content/view/18243/1/

Now if NVIDIA realized that they need to redesign this mid range part to overcome the yield and power issues that plague Fermi, then there is a chance that with high clock speeds this part could challenge the 5830 and 5850, and certainly beat up on the 5700 series cards. If they just kept doing metal layer spins like they did with Fermi, then this part is more likely to only hold it's own against the 5770 which ATI can produce much cheaper. A revised Fermi core tuned for power consumption and yields would produce a 512 sp part that could challenge the 5970, and a dual GPU card to surpass it. Of course by the time it came out it would be dealing with ATI's next chip.

As much as we like the high end stuff (yay 5870 \^_^/ ) it's the midrange and lower end parts that drive revenue. If nVidia doesn't get that part right then their bottom line will really suffer. The GT 240 is doing pretty good for them right now, but if ATI were to lower their prices across the board, then a cheap 5670 would force nVidia to lower the price of their GT 240 and likely cut into their margins even more.
March 26, 2010 7:54:13 PM

photonboy said:
If you're deciding on your stocks based on such a simplistic analysis of NVidia as a whole you clearly don't understand how to analyze stocks.

NVidia is a lot more than just how its latest DX11 offerings do in the short term.

NVidia has Tegra and Tegra 2 for netbooks, media players and even the Nintendo DS I think. They'll probably have an investment in one or more of the future consoles (2012?).

NVidia has investments in computers for scientific firms. You should read about the NVidia relationship in developing the movie Avatar.

To understand about stocks for NVidia you need to do a LOT of research on Intel, AMD/ATI, NVidia, Google and Apple.

You need to understand most of the technology, the history of technology development, future trends etc. It is complicated.

Apple, for example wasn't doing very well then they came along with the iPod which not only saved the company but paved the way to the iPhone (Apps Store) etc.

A good company is diversified without being cash and research poor. NVidia is poised quite nicely but deciding when to buy and sell Stocks is very risky in any company and not a good idea if you don't have a firm grasp on the entire Technology field that NVidia is part of.


So please enlighten me. If my analysis is correct and Nvidia dont have anything competitive for the low to high end (except Fermi for the very high end) and will loose a large part of the market, what do you predict will happen to the stock?
!