http://www.fudzilla.com/content/view/18170/1/
Printable View
Even though the benchmark numbers are slowly trickling in, I'm not fully convinced of buying a card only to be overshadowed by the end of the year with a 512 part.
I don't know, maybe the reviews will give a bit more clarity in how fast it is with 480 cores, and how quick a 512 part would be. Future proofing as much as possible is important for me, since I don't have the funds anymore to update my rig every 6 months. My G80 really has blown me away in how long it has been with me. The GFX card I've owned the longest. I'm actually not going to sell this one to buy my next card. It's too good. :D
But there you go, there is always something else right around the corner really. You can wait forever. :p:
The way i heard it Nvidia disabled the shaders and they originally had 512 SP working but with high tdp and lower freq. so they went with fewer shaders that made the card fast and also less power hungry.
In any case a 512 part will come after all the arc supports 512 after all...
First post updated
short term, no, long term, most likely yes...
hmmm under 50%... well thats not a surprise, is it?
if it would be ABOVE 50%, that would be news... but under 50%... and that doesnt really say anything, 5% is under 50% as well :D
but if its somewhere in the 40% range, then thats great news for nvidia!
and that would explain the low prices for the gf100 cards too!
Graphical view to easily spot the differences
http://www.framebuffer.com.br/sites/...0-44-41%5D.png
If we take the score posted a few posts ago (X9322), the graph changes a bit
http://www.framebuffer.com.br/sites/...0-51-09%5D.png
If the 512 part makes it in the next 3 months , it would be totally worth waiting for. Current 480 benches being impressive though only then we should know in a few days.
Ok lets do an informal survey.
WHO HERE ACTUALLY HAS AROUND $1000 TO SPEND ON VIDEO CARDS TODAY?
Does it really matter if the almighty GTX480 is 8% or 18% faster?
Look at the GTX470 - its gonna get eaten alive by HD5850 rebate promotions - a price war that would drag nVidia into bankruptcy.
hey nice for ati to treat its customers like fools wait 6 months for a driver just to show up nv lol
not that the 5xxx cards had all its use for the first two months anyway so really only 4 month head start.
6 months for cf proiles? lol yeah ati can do no wrong :rolleyes:
i hope ati they rush out a 5890 or w/e for cheaper to further show that its first adopters are fools? lol j/k
its not like ati needed to rush out the cards with crap drivrs support(5980 is still bad)for any dx11 games just to capitilise on the dx11 pay off with all the games they said would come out shortly..i think 1 or 2 dx11 games came out in six months not to mention it seems that nv dx11 card will run those apps better but will see
ATI were not sitting on the driver, if they were it wouldn't be as buggy as it obviously is. They probably released it early so it could make it into the fermi reviews, as the reviewers need it about a week before they start writing so they can do all the benchmarks needed.
For reference. I7 @ 4GHz no HT, 5870 at stock, Catalyst 10.3a,Extreme preset.
http://img686.imageshack.us/img686/6...501200c.th.jpg
I'll assume the 965 is at stock ( hard to say since PhysX skewed the cpu score )
You do realize that the GeForce product line does not constitute the majority of NVIDIA's business model or potential profits, right? They could easily promote the GTX 480 and GTX 470 as loss-leaders and still come out FAR ahead on profits from the Fermi architecture.
Why? Because proper marketing of the GeForce series will in a round about way promote the Fermi-based Quadro and Tesla-based products which are both the real money-makers. They also both occupy areas within markets where ATI has next to no presence. Add their ION and Tegra lineups to things and you'll begin to see there's absolutely no way that selling GeForce cards at a narrow profit margin will substantially hurt NVIDIA's bottom line.
Just checking this thread to see where nvidia is at these days. I'm thinking about purchasing another card... I gota just say that I'm very glad I purchased the 5870 when I did. Performance has been great across 3 displays and I still use the 4th for vent and my cpu monitoring (cpuz, realtemp, etc). So instead of switching back to nVIDIA I think I'll just go buy another 5870 in a few months.
what $1000 are you guys talking about? GTX 480 will cost $500, not 1000.
Probably talking about TWO 480, not just one :p:
It would matter to me a while back... Rather have a stable and fully compatible video card than one that doesn't give me all candy in games. But things have changed since i tried this 5970 and Catalyst 10.3a... It's so much better than what it used to be. And i am going to add a GT240 for PhysX (as i really enjoy PhysX when it's present), so all my "problems" are solved and right now i don't feel the need for nVidia cards.
Many people will prefer otherwise and die-hard fans in special :rolleyes:
As for bankrupcy... ATI would fall first, nVidia has many more sources for the €uros and Dollar$.
But even with 512PS (or Cuda Cores as they call it now) the difference in performance won't be THAT high and i bet the price will :mad:
cheers, nice graphs :toast:
i think nvidia cut the 480 down to 480sps because 512 wouldnt have made a big difference... they still wouldnt get close to the 5970... so they went for 480sps which is still enough to beat the 5870 but means they can harvest a lot more 480s and... make more money per wafer...
would be cool if nvidia allows people to unlock their 480s to MAYBE get the full 512sps... at least for the time beeing when there isnt a 512sp card for sale... but i dont think they will do it :(
pff... at best it would result in an 8% perf boost... most likely less... whats up with all you guys looking for a reason to wait? nvidia is 6 months late, the cards are STILL not for sale, h3ll they are still not even LAUNCHED at this moment, and you are already looking for reasons to wait even more? 0_o
hurt, yes... cause them losses for a quarter, two or even three... maybe... drive them into brankrupcy, definately not... nvidia stumbled with gf100, but they are still standing... the next 6-12 months will be very interesting... if ati can place a heavy punch at nvidias chin in this time frame, they have a very good chance of seeing nvidia kiss the floor.
but look at nvidias market share in gaming graphics... ati is at 30% and nvidia at 68% and i think in laptop and professional segments its similar. ati needs more than 6 or even 12 months dominance to hurt nvidia or even knock them out... even if fermis successor/shrink fails it wont be enough to drive them into bancrupcy...
that was just a speculation... i think its likely, and yes, its kinda lame... but then again, a 10% boost doesnt make a resolution or setting playable... 10% is not enough to go from unplayable to playable, its just enough to go from "not that nice" to "good"... and i dont think its just ati doing this...