Originally posted by: keysplayr2003
Hindsight is 20/20. I expected at least a Ubermensh high end refresh in November. Die shrink, core tweaks, faster memory, some feature improvements, etcetera. Damn you AMD. Damn you......... dammit.![]()
Originally posted by: keysplayr2003
Of course ronnn, I humbly apologize. It's never AMD.
Originally posted by: apoppin
told you so
AMD didn't do it "on purpose", you know
-otoh, nvidia IS [wisely] taking advantage of the situation to maximize their profit
Neither you nor i can feel really bad ... we have had capable cards since early Summer ... i feel i am getting my money's worth out of my upgrade
Originally posted by: Cookie Monster
Looks like the G92 will be named 8800GT. Priced at $199 (256mb) and $249 (512mb).
Im still wondering, G92 will have 64SPs, but if they are DP (dual precision) then this might equal to having 128 SP SP. (single precision steam processor).
I actually think nVIDIA might push the idea of Quad SLi for the second time round. (Or even 3 way SLi as the next gen nforce boards support Full 3 x16 PCI-e slots) Its just the software side things that are lacking, and doesnt make the idea feasible, well to us anyway.
Quite the potential to buy for most of us looking for budget yet high performing GPUs. RV670 vs G92. This will decide round 2 of the DX10 battle between AMD/nVIDIA. We dont lose either way since both cards look really promising!
Originally posted by: ronnn
Originally posted by: keysplayr2003
Of course ronnn, I humbly apologize. It's never AMD.
No I must be wrong, because it of course is never nvidia.
Originally posted by: keysplayr2003
Originally posted by: ronnn
Originally posted by: keysplayr2003
Of course ronnn, I humbly apologize. It's never AMD.
No I must be wrong, because it of course is never nvidia.
Instead of taking this route, why don't you try to clarify how you think AMD/ATI has nothing to do with Nvidia's business decisions. To be clearer, why did you state as fact, that AMD/ATI's lack of performance this gen has nothing to do with Nvidia's roadmap decision. These aren't your exact words, but the point you tried to make is the same.
If AMD/ATI had released an R600 that outperformed an 8800GTX, you don't think Nvidia would have to step things up and release what they probably intended to release "if" R600 had kicked the G80's buttocks? As it stands right now, Nvidia releasing a new high end GPU right now would just be kicking it's own butt. It's why G80 pricing have stagnated. Nothing to drive down prices, and no reason to release new hardware to drive down there own prices either.
If you were a GPU company and you had the most powerful product out there, then your direct competition releases something not quite as powerful, would you then rush to release something better than your already dominant GPU? Why would you? Why go through the expense of making that happen when your current lineup already takes care of business? Why wouldn't you scrap current plans and focus on what would have been your next gen intended to take on the competitors next high end?
You can't defy this logic, but I see you trying. But why?
Originally posted by: ronnn
Originally posted by: keysplayr2003
Originally posted by: ronnn
Originally posted by: keysplayr2003
Of course ronnn, I humbly apologize. It's never AMD.
No I must be wrong, because it of course is never nvidia.
Instead of taking this route, why don't you try to clarify how you think AMD/ATI has nothing to do with Nvidia's business decisions. To be clearer, why did you state as fact, that AMD/ATI's lack of performance this gen has nothing to do with Nvidia's roadmap decision. These aren't your exact words, but the point you tried to make is the same.
If AMD/ATI had released an R600 that outperformed an 8800GTX, you don't think Nvidia would have to step things up and release what they probably intended to release "if" R600 had kicked the G80's buttocks? As it stands right now, Nvidia releasing a new high end GPU right now would just be kicking it's own butt. It's why G80 pricing have stagnated. Nothing to drive down prices, and no reason to release new hardware to drive down there own prices either.
If you were a GPU company and you had the most powerful product out there, then your direct competition releases something not quite as powerful, would you then rush to release something better than your already dominant GPU? Why would you? Why go through the expense of making that happen when your current lineup already takes care of business? Why wouldn't you scrap current plans and focus on what would have been your next gen intended to take on the competitors next high end?
You can't defy this logic, but I see you trying. But why?
I already gave my reasons. If Nvidia has developed a new high end card, lots of money has already been spent. The way to recoup that money is to release the card and than those that want the best would upgrade (as they always do). Nvidia would than make money. Likely this was the reason for releasing the ultra. Not releasing a product brings in nothing, especially if the competition is readying a mid range priced card that is close to high end performance. For Nvidia to hold off, because AMD has - would mainly suggest collusion in early planning (ie a couple of years ago). I hope neither would do that. Waiting to release and giving others time to catch up would be very risky.
A mixed answer is possible also. Nvidia has developed a high end working chip, but there are production problems. Since Amd seems to have nothing there, Nvidia does not need to release a product that would either lose money (exceedingly poor yields) or create negative publicity by having issues.
I hope this moves us off this personal insult thing. :beer:
Nvidia Corp. not only plans to refresh its lineup of performance graphics accelerators this year, but also intends to introduce its 3-way SLI multi-GPU technology aiming extreme performance enthusiasts. But will the new triple SLI technology truly become a high-performance solution, or will share the destiny of Nvidia quad SLI?
According to a slide published by Expreview web-site, which is presumably from Nvidia Corp.?s roadmap, 3-way SLI is Nvidia?s new ?ultimate gaming platform?, which will offer ultimate performance in three-dimensional games. ATI, graphics product group of Advanced Micro Devices, also plans to offer 3-way CrossFire multi-GPU technology, however, one of the graphics processing units (CPUs) in such configuration will be able to compute physics effects in case the game supports this capability.
Initially Nvidia plans to enable triple SLI support for the top-of-the-range GeForce 8800 GTX and Ultra graphics cards, however, eventually it may support 3-way configurations of other GPUs as well. Systems with three graphics cores will be powered by Nvidia nForce 680i as well as nForce 780i platforms with the former supporting PCI Express 1.1/1.0a, whereas the latter featuring PCI Express 2.0 along with a special ?BR04? switch for more efficient multi-GPU operation.
Exact capabilities of 3-way SLI platforms were unclear at press time. However, the exact feature set will mostly depend on driver support. Besides, performance boosts over single- or dual-GPU configurations will also depend on drivers. Back in 2006, when Nvidia unveiled its 4-way SLI technology, actual systems featuring four GPUs could not offer leading performance in all games due to poor drivers. Currently quad SLI technology is not supported for Windows Vista and customers who paid over $1000 for graphics cards alone have to take advantage of only two GPUs instead of four.
Nvidia did not comment on the news-story.
Originally posted by: ronnn
Why would Nvidia skip a high end card, when many would happily drop their 8800 ultra for something faster? Also gives Nvidia a good driver head start to get it out there. Nvidia could charge what they liked for a something twice as fast as current. Actually the GTX has dropped in price (at least in Canada) and will drop more if the 2950pro is as rumoured.
I think you are right that the x2 card is the likely roadmap.
Amd had made their own problems and Nvidia also makes their own problems.
researching a new gpu is relatively cheap, changing over their fab production to a new gpu is NOT. They have a breakeven point which they must achieve to make a card upgrade worthwhile vs continuing to produce what they have right now. If they can charge $50 for an 8900 ultra but they only sell 1,000 of them, it's definitely not worth it for them to switch. That's the beauty of the midrange gpus; their market is many times larger than high-end so you need a much lower % of users to adopt your new card. also, midrange users have been without dx10 thus far, making it the next logical battleground. we all know that halo-effects will help all other cards sell if you have the fastest, but nvidia's fortunes can turn quickly if DAAMIT can produce a better midrange card for the same or less money.Originally posted by: ronnn
You still miss my point - if Nvidia has something faster right now - many people who already have an ultra would buy. Which is cash in hand. Assuming it will still be the fastest in 6 months is risky - versus guaranteed sales. Also would get a higher premium, as its lead over the 2900pro (soon to be 2950) would be greater. Or for the Nvidia only audience, the lead over the gts.
Actually I never said Nvidia does not have a high end because of difficulties. Just said it is possible. Personally I think they are following their road map, which appears to be very similar to amd's. Multi gpu - that are smaller, and produce less heat. Heat and power seem to be a barrier at this time.
Still there are rumours that the g92 is not the best, likely fueled by gossip sites or as you say - by Amd or Nvidia marketing. Hopefully will be interesting times this November and both will be on the shelves.
