Printable View
Showed where? Their financial statements say otherwise.Quote:
Originally Posted by saaya
Some things take a certain amount of time no matter how much money you throw at it.Quote:
well they didnt say it would solve mfc problems, it would help to get a new cut down and/or reworked fermi out asap... i dont know about that, but it makes sense... there are a lot of things you can do to cut ttm that have nothing to do with mfc, but they cost money...
As usually.Quote:
Originally Posted by Jowy Atreides
Some people are saying there will be no Fermi @ CEBIT to look at for everyone (just closed demos). Means that the launch in March is likely to be a paper launch, ouch... Not surprising we don't get much info about Fermi.
Hope CEBIT will shed some light at Fermi an its price and performance, I am really curious if all these delays are actually worth it.
Come to this thread after a few weeks. Same stuff and yet another post by Charlie lacking any factual information. I would suggest everyone remembers or bookmarks Charlie's article and brings it up again in March / April. ESPECIALLY regarding the lower-end GF100 cards not being taped out. ;)
source; http://kotaku.com/5473644/metro-2033...stuff/gallery/Quote:
, NVIDIA senior vice president of Content and Technology Tony Tamasi seems very impressed.
"Our engineers have been working closely with 4A Games to ensure that Metro 2033 takes full advantage of some of the spectacular effects made possible with the next generation of DX11 hardware. The 4A Engine is one of the most advanced game engines we've ever worked with, and with DX11 enabled, Metro 2033 is undoubtedly one of the best looking PC games of 2010. Combine this with NVIDIA 3D Vision and PhysX and you will experience mind-blowing visuals."
the game gets released March 16th...Fermi should be a week before/after the game is released if Nvidia high head is promoting the DX11 game.
^ Hope they put a free copy of that in with each fermi :P
I'd buy the game just to get the free Fermi with it.Quote:
Originally Posted by Bodkin
That list of 9 I posted was aQuote:
Originally Posted by saaya
"pick and choose adventure"
of which (if not all) faults/challenges
are causing the FERMI trouble/delay.
Personally I dont think things are apocalyptic bad. Even nVidia-die-hards deep down inside probably knew some SP would be disabled and there would be power issues.
But like R600, whether the power/yield issues are resolved or not, Fermi will launch.
CHARLIE SAYS
Is he suggesting that only a few thousand Fermi will be made, and then for 6-9months nothing until the 28nm die shrink??Quote:
The two real fixes, doubling the vias and redesigned circuits to minimize the impact of transistor variances both require a full re-layout. Both also cost time and die area. The time is at least six months from tapeout, if you recall. Fermi taped out in late July and was only slated for small numbers at retail in late November, a very unrealistic goal. A B1 spin of the chip would be at least Q3 of 2010 if it tapes out today, and it won't have a useful life before it is supplanted by the next generation of 28nm chips.
...
The die size wall is similar, you can only fit a mask of a certain size in the TSMC tools. The G200 pushed that limit, and the changes to Fermi/GF100 would likely push the chip to a size that would simply not fit in the tools needed to make it. At that point, you have to look at removing units, a task which adds even more time to the stepping. The only way to make it work is a shrink to 28nm, but the first 28nm process that is suitable is not going to have wafers running until the last few days of 2010, best case.
If AMD can get acceptable yield rates, and very good clocks, the only thing stopping nVidia from the same is a little corporate espionage :up:
they should have had to come with better answer imhoQuote:
Originally Posted by mapel110
@eric66..i agree charlie may not be 100% but i bet he's atleast on the right track..i mean for nvidia to have nothing to say with a month to go..cant be good..of course we hope for the best for all of us..its hard to expect greatness this late..
The new rocket demo from nvidia looks great, cant wait to grab one, or two of these cards.
People forget pretty quickly, G80 was pretty much the exact same thing.Quote:
Originally Posted by InCredible
+1Quote:
Originally Posted by HelixPC
Especially since it has a bit of game built in :up:
i haven't forgotten..i remember that a month before release we had details stating the core clock shader clock and memory along with the bus for certain..along with review sites already published the fact they had a look at a sample..i mean now for fermi we have a lot of specs but any published facts?Quote:
Originally Posted by Tim
just a little proof from dailytech
oct 5 and 4..g80 was released nov 6th
http://www.dailytech.com/article.aspx?newsid=4441
http://www.dailytech.com/Power+and+t...rticle4442.htm
http://www.marketwire.com/press-rele...DA-1118552.htmQuote:
Originally Posted by eric66
They have an answer.
One of the best quarters ever.
But nvidia is so screwed. Right, Charlie...
Hmm yeah, looks like the market, consumers and the big OEM's are actually rewarding NV for rebranding, proprietary tech, being 6 months late to market, etc., etc. Seems very odd to me, but I guess the people have spoken, and this is what they want the PC industry to become! :)
Charlie is in heaven. At least until he wakes up.
i think he'll be there until Nvidia finally puts up...Quote:
Originally Posted by NaMcO
Ummm, you've just described what the PC industry has been since it's inception....Quote:
Originally Posted by flippin_waffles
Such a tease. I guess March isn't too far away.Quote:
Originally Posted by SKYMTL
The lower end/mid range parts will come very quickly after high end, I heard from someone at Beyond3D.com. 1-2 months after. No delay for them. There is also no logical reason why this should happen.
March comes in 10 days and we still without any information.
Metro 2033 will be released on 19, that gives us a month, and I bet that we will not see any card to shop until last april and in low quantities.
No one said that GF100 would be released in conjunction with 2033. All Tony said was that the game would take full advantage of their next-gen architecture.
As for the lower-end parts, I am sure NVIDIA has them in the pipeline considering how scalable the architecture is. Will we hear about them right at or after launch? Only time will tell.
The high end part with 512 Cuda Cores will really be very rare and for the first months nothing more than a press edition, but the 448 Core part will be shipped in higher amounts.Quote:
Originally Posted by Kersh
lol still no official news... we might see an apri's fool joke about fermi this year LOL maybe a fake launch ??? still 1 1/2 month to go
Sounds like more games running 100% only on Nvidia+Physx :rolleyes:Quote:
Originally Posted by Monkeywoman
I did:confused: Which part? Are you refering to intel's monopoly abuse(if you are I %110 agree) or something else? I'm not sure if i'm not following what you are saying correctly, or if you aren't following what i'm saying correctly! :p:Quote:
Originally Posted by xVeinx
aint we there yet :/
i'm hoping they've been taking lessons from ATI on how to keep a secret, recently ATI has been supprising everyone at it's product launches with features and performance. anandtech's recent article on the birth of cypress gives good insight as to how they compartmentalize certian aspects of design to controll the number of people who know what's going on. also, they apprear to be using different code names for the same feature across departments to confuse anyone on the outside looking in... if nvidia has begun doing similar things, that would explain the silence... i hope.Quote:
Originally Posted by Sn0wm@n
Big LOL to the Nvidia marketing director for doing this::shocked:
A marketing director for Santa Clara-based Nvidia is facing a misdemeanor charge after he allegedly told a flight attendant at San Francisco International Airport that he had a bomb in his jacket, prosecutors said Monday...
Full Story and Source - http://www.mercurynews.com/breaking-...nclick_check=1
EPIC FAIL for Yushing Lui :ROTF:
no wonder things in nvidia are not shaping up with employees like this :shakes: , and no sign of Furbi yet.
Since Charlie still thinks the yield was 2% with A1, I'm assuming his yield numbers are still a bit off so his cost per GPU is also off.Quote:
Originally Posted by Tim
those results are tmscs favor for nvidiaQuote:
Originally Posted by mapel110
Wasn't the 2% yield myth debunked a couple of months ago? It was something lost in translation or something like that...
Correct, supposedly was ~10% but higher than 2%.Quote:
Originally Posted by annihilat0r
It was neliz who posted the info before Charlie posted his 2% article, neliz also somehow knew that Charlie would get bad info. ;)
http://forum.beyond3d.com/showpost.p...postcount=2123
too much silence is indicative of fail... even when g80 was near launch nvidia did have more info then today with fermi...Quote:
Originally Posted by 570091D
april fool here we come :D
AMD's biggest surprise was the 4800's (and mayyyyybe 3800's somewhat) but definitely not 5800's and not at all 5700's. The performance of 5800 was expected and it was even disappointing that AMD didn't take any risks. Juniper performance was flat out disappointing.
well he might be right... after all igor is the king of destroying credibility... his own that is... :DQuote:
Originally Posted by mapel110
compute shader for games? idk... except for ai and physics, what could you use it for? and nvidia will push devs to use physix instead of compute shader, so all thats left is ai... and that works very well on the cpu and nobody likes to touch ai programing cause its so damn complex...Quote:
Originally Posted by trinibwoy
damn... i cant find it... it was an nvidia press event... very recently... like a month or two ago... and they showed that they spend a lot on rnd and that it steadily increases year over year... they had a linear graph showing rnd spending...Quote:
Originally Posted by trinibwoy
i dont remember where i saw it... it was a site i never visited before... i think i got linked there from the news section here... hmmm
well nvidia DOES have a booth at cebit this year... they didnt in the previous years iirc... so i think they will "launch" it but whether it will actually be for sale after cebit... who knows...Quote:
Originally Posted by zalbard
new one? link? :)Quote:
Originally Posted by HelixPC
hmmmm competition caught up and at the same time their new architecture is delayed by at least 2 quarters if not 3 (it was supposed to come out before nov originally), their gpgpu segment fails to kick off, their tegra segment fails to kick off, they lost their chipset business almost entirely thanks to intel locking them out and amd not extending their license either, the gaming industry moves more and more away from the pc as a gaming platform and to consoles and mobile devices instead... yeah... this is one of nvidias best quarters ever... 0_oQuote:
Originally Posted by mapel110
hah, thats hilarious :DQuote:
Originally Posted by chispy
in germany there was a similar case a few months ago, where a man claimed he had a bomb in his pants... referring to his schlong... unfortunately mr testosterone insisted on this joke when asked repeatedly by security personnel, so they took him into custody, he and his family missed their flight and the airline refused to give them new tickets since it was his fault they missed the flight. he ended up losing around 4kE in ticket costs, had to pay for his own interrogation which was another few kE, plus a fine for obstructing traffic and operations of the airport of a few kE, PLUS his wife/gf supposedly left him with the kids after this little joke... :lol:
they should have kept it at 256bit :(Quote:
Originally Posted by annihilat0r
would have been such a sweet card...
Yeah, ati's biggest hit lately has been the 4800's. Nvidias has been g80/g92. I still think the GF100 will be powerful, they just don't have enough full parts at the highest speed bins to start braggin and throwin around benchmarks. GF100 is quality performance IMO. (just can't deliver them in large quantities - hehe) Conversely, the chip's design is more "quantity" for it's time. I think the "Fermi2" refresh will be a hit. This first 512 shader GF100 part should be amazing, yet hard to find.
It's awesome how people fail to see that the R600 design too was long, and the card ended up big, hot and disappointing; but AMD used nearly the exact same design in 3000s, 4000s and 5000s and all those cards kicked butt.
If this is a good design, it might fail for Nvidia this time (which I believe it will) but could be the base for some awesome products later this year or next year. An underwhelming product doesn't mean that Nvidia is dead.
This is a really good point. :up:Quote:
Originally Posted by annihilat0r
I too think Nvidia is... Well not "doomed," but in trouble for a little while. But perhaps as you said it'll be a long term gain.
Since November. bring it NvidiaQuote:
Originally Posted by ajaidev
A great deal of graphics effects can be implemented as massive parallel computing tasks, getting around the limitations of the traditional shader pipeline and repetitive texture fetching, like the various ambient occlusion methods, full-res lens effects, order-independent transparency (a bliss for all the deferred engines out there) and many more.Quote:
Originally Posted by saaya
Let's not forget that nVidia ended up dominated sales for the period between R600 and RV770 (although RV670 help stem the tide). If Fermi really is an R600 type scenario than nVidia is in far a dark time.Quote:
Originally Posted by annihilat0r
Charlie's not ready. He will need a new pair of underpants :rofl::rofl::rofl:Quote:
Originally Posted by ajaidev
No.Quote:
Originally Posted by ajaidev
Nvidia, are YOU ready?
November was soo last decade.
This is one creepy nerd on the pic! :rofl:Quote:
Originally Posted by ajaidev
Source: http://www.fudzilla.com/
Not so bad if you ask me. Higher TDP as expected, but looks like the cooler will be more decent so it's not a big deal.Quote:
We got some word that Nvidia's upcoming Fermi-based Geforce GTX 480 single GPU card should be as hot as the GTX 285. These are still preliminary results, as Nvidia still has to send out the final cards with the final coolers, but this looks like slightly better result that the GTX 480 that we saw at CES 2010.
The TDP of the GTX 480 will end up higher than on GTX 285, but with the help of a good cooler the thermals should be very close.
The noise lever should also be very comparable and the card should not end up much noisier than GTX 285. We heard numerous times that GTX 480 will end up with different cooler than the one we saw at CES 2010 and that was showed to Nvidia's loyal press at Editor Days.
All in all, you won't burn your fingers on the GTX 480, but it won't be cold either.
Last decade or not, you STILL don't need new cards or DX11.Quote:
Originally Posted by Jowy Atreides
You can already play Battleforge, BFBC2, Dirt 2 and CoP in DirectX 11 mode, though. AvP's almost ready.Quote:
Originally Posted by NaMcO
I know, but where's the differences, i mean something that really notices like Unigine's Heaven demo (which doesn't even count as a DX11 product - it's only a demo/benchmark).
I agree.Quote:
Originally Posted by NaMcO
However, some people like to be able to crank all the settings up in any case, so this would be a good purchase for them. :yepp:
I actually doQuote:
Originally Posted by NaMcO
I'd like to see the difference over the entire demo rather than just the part with heavy tesselation, as I'm not sure the picture is as clear cut as nvidia says it is.Quote:
Originally Posted by NaMcO
If tesselation doesn't impact on anything else and it's basically free polygons then fine it's superior. If however there is a impact somewhere else, that would mean games would have to lose texture or shader detail to crank the tesselation level above the 5870 comfort zone.
That's not something I'd like to see, as games have been using shaders to give objects a more realistic look for so long having a balance of both would be much better than losing one for the other.
Over-reaching uarch that is ahead of its time normally has that problem. Just look at the NV30; people lament about how a terrible GPU that was but forget that it was the basis of the NV40 and subsequent 7900 series that was wildly successful.Quote:
Originally Posted by annihilat0r
What ATi appears to have understood is that the product itself isnt what draws sales; its the execution. Eventually the power/die issues with solve themselves in time. Just make damn sure you get to market first and make sure the features and rollout is executed flawlessly and for the love of god know your limits.
Nvidia desperately needs to learn this lesson...:shakes: It is one thing if you are a gov't agency like DARPA; sure unlimited funds and unlimited time? Go for it make your Fermi. Nvidia however appears to have forgotten they are in the buisness of making money not padding their company ego or solving world peace one tesla card at a time.
Make a card the PEOPLE want; NOT what YOU want. Simple Buisness 101. ATi gets it; nVidia clearly doesnt.
Great post, this reminds me of what Microsoft did with its XBOX 360, "gets first in the market, try us", by doing that Microsoft gave the chance to customers to see what was all about.Quote:
Originally Posted by Sentential
ATI used to be so good before the X2900 with its X1900 which was the best card around, now is just like those days.
The customers want to buy, if a new standard came up then give the option to them to adhere it, my point is If Fermi is 10% faster than the HD5870 would you replace your HD5800, remember that you already have DX11? wiser buyers would not do it even if it is 15% faster.
^ I wish I could get a 5870, but I love 3d vision too much. Hope someone comes up with a ati version soon now 10.3 has basic support for it
Charlie hits again.
Nvidia's Fermi GTX480 is broken and unfixable
2-3 pages late bro :DQuote:
Originally Posted by kl0012
Charlie is entertaining :D
A1 TapeOut was ready by the end of october, so the earliest date would have been late November for a launch with Availability in December.Quote:
Originally Posted by saaya
DesktopGPUs grow about 19%
QuadroGPUs grow about 25%
Overall GPU-Buiness increased by 22%
nvidia delivers RSX for Playstation3 and PhysX for WII, Playstation3 and Xbox360.
The ION-Chipset is in the Mac-Mini
and so on and so on. There is no reason to believe, nvidia will get in trouble in the next few quarters.
nvidia is more and more developing software and thats a much safer ground than a possible little advantage for ATI to be some months early to the market with hardware. You will see when they move to Global Foundries, that it doesn't matter that much.
The "ION-chipset" how you call it is 9400M and this chipset can be found in every MacBook as well. But this might change with i3 and i5 (an update of the MacBook Pro series is overdue).
PhysX is FREE afaik. RSX was designed by NVIDIA, but I doubt they earn money with each PS3 sold.
But Tegra looks good. To bad there are already quite a few competitors in the SOC-market and (nearly) all the manufacturers that use SOCs care about is: costs. NVIDIA has to price Tegra 2 very competitive.
We will see how this all works out.
Mapel110 what you say is true; their software is and always has been excellent and their laptop chips are doing well too. However what most people are concerned by is how they have backed themselves into a corner.
#1 They dont make AMD chipsets anymore and are now at the mercy of having AMD "allow" SLI on their boards. What if they tell nvidia to sit-and-spin?
#2 They so thoroughly peeved intel off with how they were constantly raped for SLI support back in the netburst days and then forced to support the abomination that is the 780i chipset they had to pratically beg intel for *ANY* kind of sli support for lga1156
All its gonna take is one slip up; ONE gpu that doesnt get sales before people tell them to take a long walk off a short bridge
? ever heard of iz3d? :)Quote:
Originally Posted by Bodkin
its a shame though that there is no active shutter glass kit easily available in retail that isnt nvidia branded and locked to nvidia hardware... :/
1. That would Never happen with the way AMD is struggling. They can't do anything to potentially alienate any of their customer base.Quote:
Originally Posted by Sentential
2. If they did, who cares? AMD board sales are not getting any bigger and Intel dominates in that respect.
If you're going for SLI you're going for the fastest platform there is, too, and that is S1366. I don't think it would do much harm to AMD.
RSX makes nvidia money? how?Quote:
Originally Posted by mapel110
and does sony like it? will they use nvidia in their next console? i dont think so...
mac mini? isnt that EOL for over a year?
Physix is for free, so again nvidia doesnt make any money off of it whatsoever...
hows this supposed to support nvidia financially in the next few years?
they need professional graphics, rendering movies on gpus etc, plus tegra... thats what has potential... but tegra... havent seen anything with it so far, only zune hd and everybody knew it would flop even before it came out...
tegra2 and tegra3 sound interesting, but then again, this is very console like business and i dont know how much money nvidia can make out of it...
i agree on the software side though, nvidia does a lot more software wise than ati, and they are pretty good at it! im just not sure how much money you can make off of it... in the pro market, sure... but end user?
end users are pretty spoiled in regards to software... google yahoo and microsoft throw great tools at them for years... for free... and apple has even better software than nvidia and its close to free as well...
Sony doesn't earn money with selling the PS3, but nvidia does. nvidia gets paid by Sony, not by the buyers of the PS3.Quote:
Originally Posted by FischOderAal
/edit
Okay, PhysX is free, wrong information on my side.
Sony (still) has to subsidize the PS3, I won't argue that. But I think you didn't understand what I meant. While NVIDIA might've been the one that developed RSX, Sony is the owner of the design. This way Sony doesn't need to buy the chip from NVIDIA and can produce it in their own plants (does Sony still have some?) or wherever they seem fit. NVIDIA got money for it once, for the development of RSX, but not for each chip in each Playstation 3.
At least that's what I read somewhere.
IMO, IZ3D suffers from a lack of money and thus isn't able to finance development. Their technology incorporating a pair of lower-quality panels into a single display can't even hold a light to 3D Vision and other competing technologies.Quote:
Originally Posted by saaya
Tegra is cool, and I love my Zune HD as it loads songs much faster then a Ipod Touch. But Nvidia is a GPU company, thats what they do. Sure they have made chip-sets, good ones too, back in the my AMD FX Nvidia was the only chip set I would use. They no longer can make chip sets for the big markets, Ion has more or less failed due to how Intel is pricing Atom and the only point of Ion is to say you can "kinda" play games on it and maybe watch a blue-ray if you can find a dual core Ion box, which is hard to find retail here in VT (we only have best buy).
I would expect the new MS phone to use Tegra so theres some money for Nvidia. If Nvida wants to turn in to Via and make net-book and cell phone chips, go them. But that won't help the market much when it comes to GPU pricing.
Where's "Fermi Thread Part 3"? Another usless 670 posts of crap! :(
..damn 671, mine!:D
You very much underestimate the performance and features of ION/ION2 and Tegra. They are both outstanding products - leaders in their segments.Quote:
Originally Posted by CadESin
But, you are right about nVidia's focus being off. ION is at mercy of Intel's pricing whims, and thus low (lower than Intel) margin. Tegra is just starting to grow a market, and likewise a small chip with little margins.
Let not forget nVidia has no license for DMI.. ie Core i3/i5/i7. Once they replace the Core2Duos in the market, ION will be platform-less.:shocked:
nVidia built the whole "empire" on high end graphics. Consumers dont buy a $50 silicon chip, but the promise of revolutionary gaming experience delivered through a $300-$600 video card. Side projects like ION/Tegra are good for some spare change, but not the bread and butter of the business.:shakes:
CAN SOMEBODY MAKE A PICTURE OF DWAYNE JOHNSON FOR MARCH!?
Even with DX11, this game is ugly and total faily :(Quote:
Originally Posted by zalbard
Sure have, but I dont want a sub standard 2d mointerQuote:
Originally Posted by saaya
Compute shaders are already used for:Quote:
Originally Posted by saaya
Battleforge: SSAO
Stalker-CoP: SSAO, Contact Hardening shadows
Metro2033: DoF
Dirt2: SSAO
what? Another deep dive day? Awesome!!Quote:
Originally Posted by ajaidev
True thatQuote:
Originally Posted by Olivon
But its not because of the mediocre textures.
Except for a few gems like Bioshock, all games have nearly identical gameplay.
Where is the destructible terrain/objects?
How about some character morphing?
Besides HL2 (2006 yo!) and Crysis, can you think of any games that use physics for puzzles?
Name just 1 game that properly animates characters climbing uneven terrain.
Maybe bring back wity henchmen dialog from NOLF.
Too often games try to be too realistic.. too much "COD" syndrome. Crist some developers need to play QUAKE1, and check out the monsters, traps, and monster-v-monster combat. (yes, I am serious)
Creating a good single player FPS experience needs someone who has played Quake 2's awesome single player campaign about 50 times.
QFT, first finished this game aged 8 :)Quote:
Originally Posted by annihilat0r
Dual fermi on route ETA May
http://translate.google.co.in/transl...ial%26hs%3D9gF
Then again
Fermi Architecture Will Only “Hit the Pull Stride” in Q2 – CEO of Nvidia.
Mass Availability of GeForce GTX 400, Other Fermi Products Delayed Till Q2
http://www.xbitlabs.com/news/video/d...of_Nvidia.html
Thank you for posting news.. It really is appreciated.Quote:
Originally Posted by ajaidev
did they mistake the year 2010 with 2011?? WTF?Quote:
Nvidia Corp. will finally start selling its highly-anticipated GeForce GTX 400-series graphics cards as well as other products based on the code-named Fermi architecture and GF100 (NV60, G300, GT300) graphics processing unit (GPU) in the first quarter of fiscal 2011, it looks like mass availability of appropriate products is only expected in Q2 of FY 2011.
of the business year :up:Quote:
Originally Posted by Arxagelos
Q2 of Nvidia's FY2011 is May-July 2010.
"Back to school sales" LOL
The big fiasco of the doomed Larrabee syndrome.
Are you talking about A2???Quote:
Originally Posted by mapel110
Because A1 silicon was done in August...
This thread is difficult to navigate. Would it be possible to link all the important info in the OP? Then again, that would imply there is something newsworthy...
thx, i didnt know that! :toast:Quote:
Originally Posted by fellix_bg
wow i didnt know... so whats ssao?Quote:
Originally Posted by trinibwoy
in theory thats true, but nobody will buy a 10.1 card over a 11 card if they perform about the same and cost about the same... if there wouldnt be 11 entry level cards this would be correct, but there are...Quote:
Nvidia’s chief executive officer did not provide any concrete timeframes for the transition of the whole lineup to the new Fermi architecture, but said that since the owners of mainstream and entry-level graphics cards hardly demand new functionality, it is not crucial for Nvidia to update currently available “fabulous” graphics chips. In addition, the speed of the transition depends on the supply of 40nm chips by TSMC.
oh man... so fermi slipped into Q2... i thought charlie was exaggerating when he predicted that most likely fermi will actually come out in Q2... everybody laughed at him and others back in Q4 2009 when they said fermi wont be out until mid 2010...
nvidia is getting really close to their tesla deadline... they promised tesla cards in early Q2... if they cant get that done, thatll hurt their business big time...
is it just me or did jensen navigate around the question if GF104 had taped out already?
he said they will have a couple of cards, he didnt say chips... a couple of cards is 2 or more, and considering that there will be 480 and 470 plus quadro and tesla, GF100 is enough to fulfill his claim of having a couple of cards and several price points... if he didnt say anything about GF104 taping out, then its pretty safe to say it didnt happen... thats nothing you hide in front of analysts, thats something you boast about... no GF104, meh :/
http://vr-zone.com/forums/561413/nvidia-quot-fermi-quot-gpus-delayed-yet-again.html
Sounds a lot like a few thousand A3 GPU boards will be shipped to consumers as is (March),
while they re-spin again in hopes of fixing whatever is causing all the trouble.
Its just rumour, but Charlie may be right about "hope TSMC fabs perfect huge chip vs respin adding redudancy like double via to make chip even bigger".
Very tough situation indeed. But unlike G200 which went with conservative 65nm, there is no "55nm" equivalent half-node shrink.. 40nm is already that!
One way or another nVidia needs to fix/work-around 40nm issues till end of '10.
SSAO is Screen Space Ambient Occlusion. It is implemented entirely on GPU as pixel shader for games like Crysis. Its an approximation to Global illumination. It allows the ambient light sources and colored surfaces to contribute to the lighting of an object (NOT specular reflections) that may otherwise be in shadow. ie when lights are on in closed washroom, everything outside washroom door is a little less dark because of light reflecting off the walls.
nVidia has a driver option for this if I recall.
Thx ajaidev, so May/june is a deception :(Quote:
Originally Posted by ajaidev
So it is Q2 afterall... Sometimes I scare myself :D
At leat we will have some reviews in march or early april. Not to defend Nivida, but Ati also had little stocks in the first month.Quote:
At present many observers suggest that Nvidia will launch the GeForce GTX 470 and GTX 480 graphics cards in March or April, but it looks like the products will not be available in truly mass quantities right after the launch.
I hope it will be something very competitive, otherwise in 3-4month Ati wil hit very bad Nvidia with 6xxx generation.
well, 32nm would have been the half node shrink for 40nm i thought?Quote:
Originally Posted by ***Deimos***
so their plan probably was 40->32 like they did 65->55 with GT200... but tsmc had to cancel 32 because of too many issues and is going for 28nm right away, which will be a different animal to tame alltogether it seems...
ahhhh ambient occlusion... i see...
thats a neat feature! i remember the demos with the UT3 engine and the crytek stuff as well...
Well, time to buy a 5870.
Fermi mass availability delayed to May-July
http://vr-zone.com/articles/fermi-ma...o-q2/8461.html
Quote:
Nvidia's upcoming Geforce GTX 400 series was expected to release some time in March. However, Nvidia claims the GTX 400 series will be released in Q1 2011 - which runs till April 30th, 2010. Reporters sugges mass availability has been delayed till Nvidia's Q2 FY2011, which runs from May 2010 to as late as 31st July 2010.
In addition to GTX 400, Nvidia's CEO, Jen-Hsun Huang claims Nvidia is all set to "launch a couple of Fermi products" that will span Geforce, Quadro and Tesla - all by July 2010. However, there are no plans to refresh the mainstream GPUs any time in the near future.
[...]
nvidia is going to get owned by ati in the mainstream market as they have no competitors at all in the 100-200$ range even after fermi is released; what makes things worse is that they seem to prepare high-end fermis only which may open the gap in their lineup from 100-200$ to 100-300$ :eek:
i thought that fermir was late but things seem to be much worse than expected; this thing has been delayed from windows 7 launch timeframe to may = 8 months....
For practical purposes, this means Fermi will compete against AMD's Northern Islands (HD6xxx) rather than Evergreen. Northern Islands should launch a couple of months after Fermi if AMD's schedule is to be trusted.
haha so they laughed at Charlie for bs news oh common nvidia you can do better than that and by the way Nvidia try to call us when you are actually ready lol
if that happens then its VERY bad for nvidia as it will be nearly a whole gen behind :/ ATI could potentially have 28nm cards while fermi's still on 40nmQuote:
Originally Posted by Teemax
and it doesnt help that TSMC are thinking about changing to 28nm now aswell as im sure nvidia would have sorted their plans for a die shrink to be at 32nm.
we shall see.. nvidia will be fine.. but it still isnt good to be so late :/
(though tbh 6 series wont come out till Q1 2011 (OUR 2011))
It's not just you. I wondered the same... If GF104 had taped out, Jensen surely would have mentioned.Quote:
Originally Posted by saaya