NVIDIA Fermi

[Pulseezar]

http://www.anandtech.com/video/showdoc.aspx?i=3651
http://pcper.com/article.php?aid=789
http://techreport.com/articles.x/17670


NVIDIA's response to AMD.

3 Billion Transistors!

[rotNdude]

Add this one too:

http://pcper.com/article.php?aid=789

[Jesse1234]

I want to fold with one of those:).

[Baron_Fel]

This is like GT200 rev 2 lol. Another totally massive GPU.

[bes]

Im thinking ATI/AMD might get their pants blown off by this card when its out..

[GrandpaQ]

Interesting response. Looks like a good 4 - 6 months away though. No Santa for nVidia :(

[l4dr0cks123]

Holy ♥♥♥♥, 3 billion transistors! This card will destroy the HD 58xx. Too bad it will be late for Xmas (this will get ATI a lot of $).

[Bad_Motha]

NVIDIA's may come first but ATI/AMD are working on two new things, "Vision" and "Fusion". So don't count them out just yet.

[steveroger]

Holy ♥♥♥♥, 3 billion transistors! This card will destroy the HD 58xx. Too bad it will be late for Xmas (this will get ATI a lot of $).

I didn't get your first point out of the articles at all.

[DorkSterr]

And it looks a lot better than the 58** series too.

http://www.fudzilla.com/content/view/15762/1/

[Bad_Motha]

Looks the same... a double stack card with the fan blower inward and blowing hot air out the back, this is an already aged design... what u mean the NVIDIA looks better?

[DorkSterr]

Looks the same... a double stack card with the fan blower inward and blowing hot air out the back, this is an already aged design... what u mean the NVIDIA looks better?

It's shiny.

[Bad_Motha]

LOL who cares, long as fits and doesn't block connections to other things, and runs the games well... who cares!

[l4dr0cks123]

I didn't get your first point out of the articles at all.What do you mean? I didn't even read the articles if what you're trying to say is what I said is completely irrelevant to the articles.

[steveroger]

And it looks a lot better than the 58** series too.

http://www.fudzilla.com/content/view/15762/1/

ATI (http://www.jayceooi.com/wp-content/uploads/2009/09/ATI_Redeon_HD_5870.jpg)

Nvidia (http://www.pcgameshardware.com/screenshots/418x627/2008/11/Unbenannt.PNG)

:rolleyes:

However, those are just the base factory mock-ups. Here is what will be released at retail:

ATI (http://forum.codecall.net/attachments/lounge/2147d1252690505-new-batmobile-1252640620801.jpg)

Nvidia (http://i68.photobucket.com/albums/i20/greenfrogthing77/newmod.jpg)

[Pickanewnameplzkthx]

Like I said in another thread. This is why I'm waiting. It looks very promising. I might get two of these, or I may wait until the next refinement of them. Cooling is niot an issue since I already have watercooling. I waited over a year to build my i7 system, and didn't end up getting it until last November. I was sooooo glad I waited. Same thing goes for these. I'm not gonna buy until I see. Something tells me I'll again be glad I waited. I'm not looking for cheap...only the max performance I can get and it'll have to last. Whatever 2 I end up settling for, they're gonna fly, and they're gonna have to crunch through Folding like nothing else out there. Gaming performance would be obvious. It's no longer just about the gaming. It's about how much work can these systems do, and how fast will they do it. I didn't get an i7 system to sit there and look pretty and idle all day, although it is quite beautiful. I got it to do work 24/7. The gaming is just a bonus. 2 of these will have to get it done. If they don't, I'll wait until there is something that will. I don;t have much doubt that Nvidia will yet again release another winner. We shall see. To Nvidia I'll say "It pays to be a winner. Are you up for it? If so, BRING IT!!!"

[Jagged Tooth]

I so what one of these. Wish I didn't get a 285a couple months back.

[steveroger]

What do you mean? I didn't even read the articles if what you're trying to say is what I said is completely irrelevant to the articles.

The OP is about the articles. I assumed that is what you were talking about. The articles said that they didn't know what the impact would be regarding the 5870. They freely stated that it was a lot of speculation because Nvidia was withholding a lot information. But the information was pretty intriguing.

[Combat Medic]

When will the nvidia Hammer Legion Members get this though there head?

THIS IS NOT THE GTX3XX OR EVEN CLOSE TO IT. THIS IS THE TESLA. IT IS NOT EVEN TECHNICALLY A GRAPHICS CARD.THESE WILL RETAIL FOR $3000+ A PEACE.

[Baron_Fel]

No, this is the graphics card as well.

Telsa and Geforce are the same hardware wise, except for VRAM.

[Bad_Motha]

Then what is this picture? Has no display outputs... wtf is that card?
http://www.pcgameshardware.com/screenshots/418x627/2008/11/Unbenannt.PNG

[Combat Medic]

No, this is the graphics card as well.

Telsa and Geforce are the same hardware wise, except for VRAM.

You are mistaken.Geforce and Quadro are the same except for Vram.

[Combat Medic]

Then what is this picture? Has no display outputs... wtf is that card?
http://www.pcgameshardware.com/screenshots/418x627/2008/11/Unbenannt.PNG

ITS NOT A VIDEO CARD.NOT ALL COMPUTER HARDWARE IS MEANT TO PLAY GAMES ON.

[Baron_Fel]

It doesnt have outputs because it doesnt need to output anything. Itll just be stashed in a server where itll crunch numbers.

[steveroger]

When will the nvidia Hammer Legion Members get this though there head?

THIS IS NOT THE GTX3XX OR EVEN CLOSE TO IT. THIS IS THE TESLA. IT IS NOT EVEN TECHNICALLY A GRAPHICS CARD.THESE WILL RETAIL FOR $3000+ A PEACE.

Sure it is. LOL.

No, this is the graphics card as well.

Tesla and Geforce are the same hardware wise, except for VRAM.

Thanks.

Then what is this picture? Has no display outputs... wtf is that card?
http://www.pcgameshardware.com/screenshots/418x627/2008/11/Unbenannt.PNG

LOL. The card that Jensen held up in the picture is just a Tesla with Fermi GPU. I just used that Tesla picture because it was a close up of a card that is pretty much identical to one Jensen held up. I was just skewing the situation to make the cards (5870 and Fermi/Tesla) look more similar. There is one difference between the two, the one Jensen held up has a shiny plate.

The whole thing was just a joke though. I was pointing out that the new Nvidia and ATI cards prior to being re-branded were substantially the same.

Once the Tesla with the Fermi GPU is sent over to BFG for example it will look pretty different because they will put their own skin on it. Just like the 5870 that goes to XFX has a unique skin.

I thought the guy who was jazzed about how the Fermi(Tesla) card look better than the 5870 was being ridiculous.

Here are some more shots, here (http://www.bit-tech.net/news/hardware/2009/09/30/first-fermi-card-pictured/1) and here (http://www.xtremesystems.org/forums/showthread.php?t=235658&page=4).

If you look close the card Jensen held (http://images.bit-tech.net/news_images/2009/09/first-fermi-card-pictured/fermi1.jpg) up at the conference says TESLA on it. It is even more obvious here (http://www.tgdaily.com/content/view/44146/135/).

I am excited about both cards. And just to prove it. ATI has a ridiculous Pig card (http://legitreviews.com/images/news/2007/pig.jpg).

[IcarusNine]

Shiny is sexy. I don't know why it took them so long to add gloss. How many more years must I wait before video cards come with a chrome grill?

[Beer_N_Cookies]

When will the nvidia Hammer Legion Members get this though there head?

THIS IS NOT THE GTX3XX OR EVEN CLOSE TO IT. THIS IS THE TESLA. IT IS NOT EVEN TECHNICALLY A GRAPHICS CARD.THESE WILL RETAIL FOR $3000+ A PEACE.

More like $700. But then again this card won't disappoint me like 2x 4870x2 did in crossfire. That was the biggest waste of money.

[Scholzpdx]

By that time ATI will have cut each 58x0 part by approx $100 if this card is good at all.

I've heard it having a paper launch but not really coming out til feb-march.

[renegadeafk]

This is a gpgpu card NOT the gt300 this is inaccurate. The guys saying its TESLA are correct.

[GrandpaQ]

I'm keeping a close eye on my NVDA stock -3% since yesterday, not a big drop but right after the "paper release" it scares me a bit. They will miss Christmas sales with a new must have vidio card and I heard they lost the Apple laptop account due to a bug in a mobile chip. I was wondering why the CEO was selling a lot of his shares, makes sense now.

Anyone else here own NVDA and concerned?

I think this new card sounds like a monster but as my wife who owns a small computer hardware company told me last night "if I can't sell one to my customers today, why would I care?" I can only think most people looking at a new or upgraded system won't be waiting for next year and will ask the sales rep what is the "best" and the answer will be the ATI 58** until this card gets released.

Bottom line - I was hoping for a faster release date and an actual consumer proto type with some idea of how they will be the leader in 2009.

Cheers

[Buff]

However, those are just the base factory mock-ups. Here is what will be released at retail:

ATI (http://forum.codecall.net/attachments/lounge/2147d1252690505-new-batmobile-1252640620801.jpg)

Nvidia (http://i68.photobucket.com/albums/i20/greenfrogthing77/newmod.jpg)

lololololol :D

[Toto pectore]

Nvidia presented fake card? Look at this picture: http://img57.imageshack.us/img57/2586/nvidiafakecard.jpg

There is 6-pin power connector on longer side and 8-pin connector on short side, but on PCB there is only soldered joints just for 8-pin (in bad place and in bad angle - turned in 90°). Also there is hole on PCB for screw but in place where 6-pin power connector is...

[Baron_Fel]

Telsa Geforce and Quadro are all based off the same damn GPU. If you discuss the architecture of one you discuss them all.

They will all have 512 cores. They will all have a 384-bit bus and GDDR5.

[Toto pectore]

A Few Thoughts on Nvidia’s Fermi (http://www.jasoncross.org/2009/09/30/a-few-thoughts-on-nvidias-fermi/) (AMD is callin it Furby :P )

[Baron_Fel]

A Few Thoughts on Nvidia’s Fermi (http://www.jasoncross.org/2009/09/30/a-few-thoughts-on-nvidias-fermi/) (AMD is callin it Furby :P )

meh, the chip has 512 cores/shaders/sps as opposed to the mere 240 in GT200. Thats over double, and there will be tons more memory bandwidth with GDDR5 and a 384 bit bus. I dont care what they did to optimize for GPGPU, this thing is going to be hella fast.

[Toto pectore]

It will, but the questions are: when and how much will it cost?

[GrandpaQ]

A Few Thoughts on Nvidia’s Fermi (http://www.jasoncross.org/2009/09/30/a-few-thoughts-on-nvidias-fermi/) (AMD is callin it Furby :P )

I hope it's not as bad as Jason Cross makes it out to be in his blog. He did raise some valid points about potential higher costs and the shift from a gaming video card leader to super computing chip vendor.

[BS_1]

I'm going to "guess" and say that Fermi based chips won't be out until February. ;)

[Batuhan27]

Interesting architecture. No doubt that it will outperform Radeon 5800. But big questions remain: cost and performance per watt. Also, ATi will probably release improved revisions of 5800 chips by the time GT300 hits the market.

[Gigi]

meh, the chip has 512 cores/shaders/sps as opposed to the mere 240 in GT200. Thats over double, and there will be tons more memory bandwidth with GDDR5 and a 384 bit bus. I dont care what they did to optimize for GPGPU, this thing is going to be hella fast.

lol, and you're calling anyone Nvidia fanboy :D while not realizing that you're one as well !

[Pickanewnameplzkthx]

lol, and you're calling anyone Nvidia Hammer Legion Member :D while not realizing that you're one as well !


You're way off base. I can speak first hand for Baron definitely NOT being a "We all know what a Legion member is". In fact, he's called me down for getting out there just a bit basically becasue I only speak about things I use and have first hand experience with.

I have seen him post positive things for both sides, and on this very subforum, there are threads about the new ATi cards where he has recommended them.

There is nothing at all "legion memberish" about his post at all. He is 100% correct. If those rumors hold true, there will be nothing close to that, and it will be indeed be "hella fast" as he posted.

If my suspicions are correct, as well as others I have talked on other forums, then this thing could very well be just that. I have alot of expeeince with Nvidia products. Enough to know that they are always planning big releases, and especially so around the Christmas timeframe. They are also very secretive, and they keep the the people they support under very strict NDA. Make no mistake though, they will occasionally "leak" a tidbit out there, and it's not by accident that they do this. It is controlled. They are in the buisness to compete and they protect their investments.

They have done a good job of baiting AMD into childish comments again. I have to admit I did laugh at that. It was kinda cute, and I also belive that Nvidia is laughing as well. They know that ATi has the benefit of the "First One Out" press, but there is no battle currently. There is no competition yet. There will be when Nvidia releases this thing, and then we'll have to see who has the last laugh. I'm not betting either way. I just sit back and watch. I'm not even making any bets this go around.

I do hope that the end product ends up being what is rumored. It might nopt be great for everybody, but for me, it would be exceptional, and the price would matter not. I've got a system that could utilize every bit of power like that, and the work that I do would take a fraction of the time it does now. I'm gonna wait an see though before I pass any judgements. Things could change between now and then, or these rumors may not be entirely correct.

[LordEC911]

There is nothing at all "legion memberish" about his post at all. He is 100% correct. If those rumors hold true, there will be nothing close to that, and it will be indeed be "hella fast" as he posted.

If my suspicions are correct, as well as others I have talked on other forums, then this thing could very well be just that. I have alot of expeeince with Nvidia products. Enough to know that they are always planning big releases, and especially so around the Christmas timeframe. They are also very secretive, and they keep the the people they support under very strict NDA. Make no mistake though, they will occasionally "leak" a tidbit out there, and it's not by accident that they do this. It is controlled. They are in the buisness to compete and they protect their investments.

They have done a good job of baiting AMD into childish comments again. I have to admit I did laugh at that. It was kinda cute, and I also belive that Nvidia is laughing as well. They know that ATi has the benefit of the "First One Out" press, but there is no battle currently. There is no competition yet. There will be when Nvidia releases this thing, and then we'll have to see who has the last laugh. I'm not betting either way. I just sit back and watch. I'm not even making any bets this go around.

WTF?
They reworked the architecture, so there is very little to base your so called "performance estimates" of "hella fast" on. Sure, they doubled the shaders and hopefully made them more efficient but with so much focus on GPGPU or CGPU, as they like to call it now, I wouldn't put it past Nvidia to sacrifice gaming performance for more CGPU performance.

Also there are other bottlenecks with their architecture that could limit their performance. The main one is clockspeeds. One source is stating that their clockspeed targets are 750mhz and they aren't even close to hitting that and would be lucky to launch GF100 @ G200 clocks. The other is ROP performance, 48ROPs at the low Nvidia core clocks is very close to 5870's ROPs throughput, plus the fact that Nvidia needs to improve their AA performance impact.

Not sure what you are talking about in regards to AMD/ATi's "childish comments." Any direct quotes for me?

Also Nvidia being secretive? If anything this reminds me of the FX fiasco of making a website, then denying they had any affiliation with the website, then a few months after paper launching, then a few months after that having cards hit shelves and reviewers and the product proceeds to get ripped apart.

Also info about G300/GF100/Fermi have been leaking left and right. When Nvidia PR has to go to Fudo and Theo to counter another websites article, you know they are more than a little upset.

Nvidia is not happy at all with their current situation, they definitely have time to improve it but there isn't much room for error, especially if they hope to launch their product in Q1 '10.

[GrandpaQ]

It looks like the Fermi card may have been faked. Story Here (http://www.semiaccurate.com/2009/10/01/nvidia-fakes-fermi-boards-gtc/)

"WHAT DO YOU DO when you have a major conference planned to introduce a card, but you don't have a card? You fake it, and Nvidia did just that.

In a really pathetic display, Nvidia actually faked the introduction of its latest video card, because it simply doesn't have boards to show. Why? Because it didn't get enough parts to properly bring them up, much less make demo boards. Why do we say they are faked? If you look at the pictures, it is painfully obvious that Fermi cards don't exist. Well, painful if you happen to be Dear Leader who waved fakes around and hopes to get away with it, but hilarious if you are anyone not working at Nvidia."

[Scholzpdx]

It looks like the Fermi card may have been faked. Story Here (http://www.semiaccurate.com/2009/10/01/nvidia-fakes-fermi-boards-gtc/)

"WHAT DO YOU DO when you have a major conference planned to introduce a card, but you don't have a card? You fake it, and Nvidia did just that.

In a really pathetic display, Nvidia actually faked the introduction of its latest video card, because it simply doesn't have boards to show. Why? Because it didn't get enough parts to properly bring them up, much less make demo boards. Why do we say they are faked? If you look at the pictures, it is painfully obvious that Fermi cards don't exist. Well, painful if you happen to be Dear Leader who waved fakes around and hopes to get away with it, but hilarious if you are anyone not working at Nvidia."

I was about to post this.

If I wasn't still ticked about the fake 9600GT's, I'd be ticked about this.

[steveroger]

Nvidia presented fake card? Look at this picture: http://img57.imageshack.us/img57/2586/nvidiafakecard.jpg

There is 6-pin power connector on longer side and 8-pin connector on short side, but on PCB there is only soldered joints just for 8-pin (in bad place and in bad angle - turned in 90°). Also there is hole on PCB for screw but in place where 6-pin power connector is...

You called it. Nice scoop.

[Pickanewnameplzkthx]

WTF?
They reworked the architecture, so there is very little to base your so called "performance estimates" of "hella fast" on. Sure, they doubled the shaders and hopefully made them more efficient but with so much focus on GPGPU or CGPU, as they like to call it now, I wouldn't put it past Nvidia to sacrifice gaming performance for more CGPU performance.

Also there are other bottlenecks with their architecture that could limit their performance. The main one is clockspeeds. One source is stating that their clockspeed targets are 750mhz and they aren't even close to hitting that and would be lucky to launch GF100 @ G200 clocks. The other is ROP performance, 48ROPs at the low Nvidia core clocks is very close to 5870's ROPs throughput, plus the fact that Nvidia needs to improve their AA performance impact.

Not sure what you are talking about in regards to AMD/ATi's "childish comments." Any direct quotes for me?

Also Nvidia being secretive? If anything this reminds me of the FX fiasco of making a website, then denying they had any affiliation with the website, then a few months after paper launching, then a few months after that having cards hit shelves and reviewers and the product proceeds to get ripped apart.

Also info about G300/GF100/Fermi have been leaking left and right. When Nvidia PR has to go to Fudo and Theo to counter another websites article, you know they are more than a little upset.

Nvidia is not happy at all with their current situation, they definitely have time to improve it but there isn't much room for error, especially if they hope to launch their product in Q1 '10.

What clocks? There are no clocks for this card yet. Everything you based your performance estimates on is pure Fud. You even included such sites as Fud to back up your claims? A rumor mill site proven over an over again to post false info and that none of the Manufacturers will get near? Then, you imply that Nvidia would have anything to do with the place and act like anything that palce would post might have a hint of truth to it? hahahahaha Seriously? Please tell me this post was a joke.

You act like you know something about these cards that nobody else does, and reply to a post of mine where I stated only that I agreed with another poster that these cards will be fast, and they will IF there is any truth to those rumors, and that is all they are. Anybody who would think that a card with over 3 Billion Transitors in it is gonna be slow really has no clue. You can spin that alot of ways, but you cannot spin it to show any other card that has that count.

As for the childish comments of AMD I belive that "Furby" comment posted above was pretty childish, and pretty typical for them, and no I don't have any direct quotes from you. Do you work for AMD? I was speaking of AMD directly, so unless you are working for them and made that comment I don't believe I do.

The fact is Nvidia is very secretive. If you have any evidence (other than some Fud) to the contrary please do post it, and post the Source of that official info, because so far, I haven't been able to find any other than the existence of these cards.

[Combat Medic]

I haven't been able to find any other than the existence of these cards.

Actually there is evidence that there are no working cards.

[rotNdude]

It looks like the Fermi card may have been faked. Story Here (http://www.semiaccurate.com/2009/10/01/nvidia-fakes-fermi-boards-gtc/)

"WHAT DO YOU DO when you have a major conference planned to introduce a card, but you don't have a card? You fake it, and Nvidia did just that.

In a really pathetic display, Nvidia actually faked the introduction of its latest video card, because it simply doesn't have boards to show. Why? Because it didn't get enough parts to properly bring them up, much less make demo boards. Why do we say they are faked? If you look at the pictures, it is painfully obvious that Fermi cards don't exist. Well, painful if you happen to be Dear Leader who waved fakes around and hopes to get away with it, but hilarious if you are anyone not working at Nvidia."

I was about to post this.

If I wasn't still ticked about the fake 9600GT's, I'd be ticked about this.

I have a feeling that was a pre-release skeleton of how the card will look. If they had an actual working card, I'm sure they would have given a grand demonstration of what it could do. The main thing with these reviews is they are releasing information to combat the AMD/ATI release of their new card and make some folks think about buying now or waiting a bit. All of us tech geeks looking for the latest and greatest may think twice about investing our money now on the new ATI card or waiting a while longer to see how this card actually stacks up. I know I'm waiting a bit longer.

[bes]

All of us tech geeks looking for the latest and greatest may think twice about investing our money now on the new ATI card or waiting a while longer to see how this card actually stacks up. I know I'm waiting a bit longer.

Purchase reviewers are giving the ATI cards high marks and also claiming 16K or higher 3D Mark Vantage scores with the HD5870.. and claim to get maxed out settings with every game...including Crysis..

Even if Nvidia releases a better card, games currently will have little use for its features vs ATI's card..

From all the things posted in links on these forums the Fermi might not perform as well as people think it will due to the way the chip is made..

[IcarusNine]

It's not that great in Crysis. It's as respectable as the dual-gpu cards, but they're talking more 30 fps than 60 fps.

As for the childish comments of AMD I belive that "Furby" comment posted above was pretty childish, and pretty typical for them, and no I don't have any direct quotes from you.

Guess who coined "Laughabee"?
Hint: Shiny pre-release skeletons are kinda his thing.

Anyway, if the GTX380 manages to double the performance of the GTX285–which is boldly optimistic–I would expect it to give 50 fps to the 5870's 30 fps on better days. However, its improvement is probably closer to 40-60%, as much as the 5870 gained on the 4890.

[Gigi]

It looks like the Fermi card may have been faked. Story Here (http://www.semiaccurate.com/2009/10/01/nvidia-fakes-fermi-boards-gtc/)

"WHAT DO YOU DO when you have a major conference planned to introduce a card, but you don't have a card? You fake it, and Nvidia did just that.

In a really pathetic display, Nvidia actually faked the introduction of its latest video card, because it simply doesn't have boards to show. Why? Because it didn't get enough parts to properly bring them up, much less make demo boards. Why do we say they are faked? If you look at the pictures, it is painfully obvious that Fermi cards don't exist. Well, painful if you happen to be Dear Leader who waved fakes around and hopes to get away with it, but hilarious if you are anyone not working at Nvidia."

OMG :D I just noticed that too! LOL at Nvidia and their 'real' fermi cards:D

[Scholzpdx]

I have a feeling that was a pre-release skeleton of how the card will look. If they had an actual working card, I'm sure they would have given a grand demonstration of what it could do. The main thing with these reviews is they are releasing information to combat the AMD/ATI release of their new card and make some folks think about buying now or waiting a bit. All of us tech geeks looking for the latest and greatest may think twice about investing our money now on the new ATI card or waiting a while longer to see how this card actually stacks up. I know I'm waiting a bit longer.

They did give a presentation of it right after he held it up. It was completely fabricated. Not true//deceitful.

With their TWIMTBP screwing up games for ATI users, fake proprietary standards (Physx, which runs great on 58xx cards/i7 CPUs) and their lack of actual working next gen hardware, nVidia looks to be PR's worst nightmare at the moment. I wouldn't take their job if they paid me 3x what I'm making now.

[Pickanewnameplzkthx]

They did give a presentation of it right after he held it up. It was completely fabricated. Not true//deceitful.

With their TWIMTBP screwing up games for ATI users, fake proprietary standards (Physx, which runs great on 58xx cards/i7 CPUs) and their lack of actual working next gen hardware, nVidia looks to be PR's worst nightmare at the moment. I wouldn't take their job if they paid me 3x what I'm making now.


Has anybody ever heard of Show pieces before? Ever seen hamburgers on a sign in a fast food joint? They're not real! They're plastic. No hamburger ever looks like that. Seriosuly man, they're plastic. They're made to look good, but would taste horrible.

This ain;t a fake card. It's an advertisement man.

I hear these cries about PhysX, and how "The Way It's Meant To Be Played Games Hurt ATi users" all the time. Who's fault is that? Nvidia didn't buy you a less capable ATi card. If you want to take advnatage of the technology you gotta buy the hardware. Now, saying that, you don't have to, but if you don't, you don't get to cry about it afterwards. That's up to the user. You gotta research what you buy. There will be propreitary standards. That's what these buisnesses are in the buisness to do. They make their hardware attractive so people will buy it. It's just smart buisness. Do you think that Nvidia cares in the least about users who are not even using their products? Why would they? What would be in it for them?

[Bad_Motha]

Like any "NEW" product lineup, it's best to not jump on it, even if it hits the market in a fast-paced manner. Let it be thoroughly tested and reviewed before buying to avoid a "bad purchase". As some of you might just jump on the ATI 58xx series and then kick yourself when the new NVIDIA cards come out. I won't say I haven't had to kick myself in the past for jumping on something, especially when it comes to a computer upgrade. So sit back and let the web-articles, writeup, benchmarks, reviews, etc. come out and then make your choice. Not to mention waiting a bit will get you a cheaper price.

[hahnchen]

Nvidia Admits Showing Dummy Fermi Card at GTC, Claims First Graphics Cards on Track for Q4 2009. (http://xbitlabs.com/news/video/display/20091002130844_Nvidia_Admits_Showing_Dummy_Fermi_C ard_at_GTC_Claims_First_Graphics_Cards_on_Track_fo r_Q4_2009.html)

NVidia confirming the above.

[Bad_Motha]

LOL, yeah it's not the first time NVIDIA has done this to keep their fans on edge and ready, even if that means letting them down by making them wait months and months for a product. But I suppose alot of them do that, i.e. ATI/AMD/Intel

[Pickanewnameplzkthx]

There we go! That should quell anymore "fake" posts. The demo is right there, and the G300 GPU has been tested, and is on track for a Q4 '09 release as has always been planned.

See? This thread is a prime example of why Fud should be kept off the Internet completely. They LIE! They do it all the time. That's all they have is guesses and lies becasue the Mnaufacturers will not have anything to do with them. They've been cutoff long ago, becasue the ring leader of that site got cutoff becasue he could not keep his word and keep things that were under NDA confidential. They have no official sources, yet they post they do all the time. They did it to Intel and made up this Huge Conspiracy that Intel was having issues with their DDR3 Memory Controllers and motherboard Makers were having fits with it when the i7's were coming up for release last year. It was a complete fabrication. No such thing ever happened, the Mothboard makers came out with boards on release day that are still the best there is out there, and there were no memory controller issues whatsoever. The launch of that entiore platform was flawless.

They did the very same thing with AMD. They told people that Phenom II's had been tested and were running 5GHz on AIR! You know? If you're gonna lie at least make it believeable. This was a preposterous lie, and everybody i knew was laughing their tails off at that. It was sooooo rediculous. No such thing ever took place, and again, it was a complete LIE! OTOH AMD did post a rediculous grpaghic of this and put it up on their website. It was an incident of the PR department on the left hand, not knowing what the right hand was doing which was the people actually testing them. I heard about all of this stuff on another forum.

I NEVER click a link to that God forsaken place. Anybody who does click a link to it puts ad money into his pocket so he can lie some more. It is an example of a site that's sole purpose is to make the Internet dumber, and it does a great job of that.

[steveroger]

I really don't care about Fudzilla and their rep to me the issue has nothing to do with whether or not the Fermi chip or it's specifications and capabilities demonstrated at the press conference. I bet those figures are true but the card held up was a fake.

The truth is that Jensen lied or misled us when he stated in the video this "puppy here is fermi" and it wasn't. The Fermi chip was not in it. It was just a non-functioning engineering demo.

Nvidia themselves verifies that it was a fake:

“While Nvidia admits that the board on the public display was an “engineering sample” and did not feature its new Fermi-G300 graphics chip (also known as GT300, NV60), the company stresses that it did show actual performance numbers that the new GPU can achieve without using simulation as well as showcased a demo that was run by its new graphics processor."

Therefore the card held up meets webster's definition of fake: "one that is not what it purports to be: as a : a worthless imitation passed off as genuine."

It is completely fair to call it a fake. Around the net in forums and website articles the lead story about Fermi is that the card shown was a Fake (http://www.overclock3d.net/news.php?/gpu_displays/nvidia_admit_fermi_was_fake/1).

Does the chip exist? Probably. They probably have the GPU running in some form on perhaps a specialized motherboard--we just don't know. But we do know they don't have a properly working demo board that they felt can be shown to the public--so they faked one.

Nividia could have just released the demo without the faked card. Why didn't they? Publicity. Showmanship. Marketing? I suppose they felt that the thunder would have been lost if Jensen said, "This puppy is a mock up of Fermi and does not contain the Fermi-G300 graphics chip, however, the demo that we are showing is driven by a Fermi-based Tesla chip."

Because of Nvidia's stupid decision to fake a working Fermin card everyone is ignoring the GPU, and questioning whether they will actually be able to release a working card before 2010. Their stupidity results in a loss of credibility about the GPU itself.

Obviously this kind of showmanship has been done before. It is always a bad idea, no matter who is doing it. This kind of thing happens with all sorts of products and the once the skullduggery is revealed backlash ensues. When will any of these companies learn.

[Pickanewnameplzkthx]

There is no probably about it. There is a real picture of it on Xbit labs, and it has already been tested. You can't fake something that exists. It's as simple as that. The Fud spread like wildfire becasue the phanboi's lapped it up (like they always do) and Nvidia posted a retraction article along with the fact that the card has been tested to shut them up. Case closed. It will be here Q4 09 like was planned. Nothing has changed. I'll be patiently waiting.

Aw, and BTW, "everyone" is not ignoring the GPU. I'm not, and neither are others in this thread. You hear this talk all the time, but it will have zero effect on the product at launch. Phanboi cries are mute. Serious enthusiasts ignore them, and they watch both companies release and then make a decision.

[steveroger]

Actually, no it was faked at the press conference and Nvidia admits it plain and simple. No amount of spin changes that. You spend so much effort trying to avoid the truth. Fudzilla isn't responsible for the negative press, Nvidia is. Fudzilla's credibility isn't at issue it is Nvidia's is. A member of our own forum recognized the fake without Fudzilla's help:

Nvidia presented fake card? Look at this picture: http://img57.imageshack.us/img57/2586/nvidiafakecard.jpgThere is 6-pin power connector on longer side and 8-pin connector on short side, but on PCB there is only soldered joints just for 8-pin (in bad place and in bad angle - turned in 90°). Also there is hole on PCB for screw but in place where 6-pin power connector is...

It is silly to argue about this. It was a dumb decision to use the mock up and pass it off as real. No one would be talking about the fake Fermi and Nvidia would not have had to issue a statement confirming that they faked it.

LOL, "Nvidia posted a retraction article along with the fact that the card has been tested to shut them up. Case closed. It will be here Q4 09 like was planned. Nothing has changed. I'll be patiently waiting."

And you say, "phanbois lapped up" the fake Fermi story? That is the pot calling the kettle black.

You can't even see that it would have a lot better if Nvidia didn't put themselves into a position of having to post a retraction.

When I say everybody is focusing on the fake Fermi instead of the GPU does include you. Here you are posting about it over and over.

Tons of us who are neutral to AMD and Nvidia products took interest in the story. Silly to claim that makes one a fan phonbio.

[hahnchen]

There is no probably about it. There is a real picture of it on Xbit labs.

No, all images of the card so far are fakes. They're just shells.

Calling out when a shell is presented as real, is not FUD.

[Toto pectore]

Anyway, does anybody know what is true about information that tesselation is not supported by hardware, but will be emulated (via CUDA?) in Fermi cards?

[Baron_Fel]

It will tesselate as per the dx 11 spec.

[Toto pectore]

Sure, but that didn't answered my question.

[Baron_Fel]

Sure, but that didn't answered my question.

I meant that however it does it, it will probably do it well because its required by DX11.

[IcarusNine]

It's probably a little bit late for Nvidia to convince Microsoft to start making a DX11.0a with less functionality* than the DX11.0b their competitor uses. =P

*Actually more(-ish?) functionality, but less effective.

[steveroger]

It's probably a little bit late for Nvidia to convince Microsoft to start making a DX11.0a with less functionality* than the DX11.0b their competitor uses. =P

*Actually more(-ish?) functionality, but less effective.

What are you talking about?

[IcarusNine]

Geforce FX series (http://en.wikipedia.org/wiki/GeForce_FX_Series#Overall_performance).

They fixed that problem with their 6000 series by jumping on DX9.0c faster than lightning, while Ati dawdled. Still, it's a painful memory of theirs that's always fun to poke at.

[steveroger]

I know about the Valve issue and the FX 5950 first hand. But I can tell you that it really wasn't as bad visually as the stories about make it out to be.

Anyway, I doubt that the FX fiasco will be duplicated in anyway.

[Baron_Fel]

What are you talking about?

Hes referencing how nvidia handled DX 10.1

[steveroger]

Hes referencing how nvidia handled DX 10.1

My last post had to do with his last post below:


Geforce FX series (http://en.wikipedia.org/wiki/GeForce_FX_Series#Overall_performance).

They fixed that problem with their 6000 series by jumping on DX9.0c faster than lightning, while Ati dawdled. Still, it's a painful memory of theirs that's always fun to poke at.

So, I don't know what you are referring to..:D

[Scholzpdx]

Has anybody ever heard of Show pieces before? Ever seen hamburgers on a sign in a fast food joint? They're not real! They're plastic. No hamburger ever looks like that. Seriosuly man, they're plastic. They're made to look good, but would taste horrible.

This ain;t a fake card. It's an advertisement man.

I hear these cries about PhysX, and how "The Way It's Meant To Be Played Games Hurt ATi users" all the time. Who's fault is that? Nvidia didn't buy you a less capable ATi card. If you want to take advnatage of the technology you gotta buy the hardware. Now, saying that, you don't have to, but if you don't, you don't get to cry about it afterwards. That's up to the user. You gotta research what you buy. There will be propreitary standards. That's what these buisnesses are in the buisness to do. They make their hardware attractive so people will buy it. It's just smart buisness. Do you think that Nvidia cares in the least about users who are not even using their products? Why would they? What would be in it for them?

When they specifically state that it is being ran in real time on a real card, it WAS faked. I don't give a crap if they actually do have hardware. Their demo was fake.

Physx is a fake proprietary standard. It runs faster on an i7 CPU anyway. It also bogs the ♥♥♥♥ out of a single GPU as it's memory intensive. Taking the 896mb GPU into a 300MB GPU in no time. I really don't care about this, but you keep hearing this from nVidiot fanboys, it gets old real fast.

TWIMTBP is fine as it is, but nVidia is PAYING developers to lock out ATI hardware. Anti-competitive practices. Batman AA is the first of many. Anti-Aliasing is a native feature on the Unreal Engine, yet nVidia paid for a lockout on ATI hardware. There is no AA for ATI cards, yet a simple vendor ID (to an nVidia part number) change on an ATI card unlocks all AA formats. You can force FSAA is CCC, but it's far less efficient than MSAA in the game.

Purchasing an nVidia card will only further all of this BS.

[BS_1]

Ever seen hamburgers on a sign in a fast food joint? They're not real! They're plastic. No hamburger ever looks like that. Seriosuly man, they're plastic. They're made to look good, but would taste horrible.
Not sure if they changed it, but we actually used toothpicks when I had to. :)

[Baron_Fel]

When they specifically state that it is being ran in real time on a real card, it WAS faked. I don't give a crap if they actually do have hardware. Their demo was fake.


They said the demo was running off real hardware, its just that the card Jensen was holding up was fake.

as for batman AA... (http://www.bit-tech.net/news/hardware/2009/10/03/nvidia-dismisses-amd-s-batman-accusations/1)

McNaughton complained that Batman: Arkham Asylum has an anti-aliasing mode on Nvidia hardware, which disappears when an ATI Radeon is recognised as the primary GPU in the system. The game also implements Nvidia's PhysX technology, too.

However, he neglected to mention that Batman: AA is based on Unreal Engine 3, which uses a deferred renderer on DirectX 9.0. Deferred renders don't support MSAA in DirectX 9.0 without a driver workaround, which is exactly what Nvidia's DevTech team helped to implement (and test). Because of the tight development schedule though, this couldn't be tested on ATI's Radeon graphics cards.

Ashu Rege, director of Nvidia's DevTech team, said that "you have no idea how tight the schedule was on Batman." He explained that there was less than a week to fix several critical bugs in the physics effects before the game went off for GfW approval because the Nvidia engineer working with developer Rock Steady on the PhysX implementation was just about to go on holiday when the bugs came to light.


So, I don't know what you are referring to..:D



What I said is actually a current example of what he was talking about :D

[Scholzpdx]

The Unreal Engine supports AA. A few mods have it implemented in all video modes, so that point is null.

At first DX9 didn't support AA, but it began to long before Batman. DX 10 has complete, native support for AA, yet is still locked out on ATI cards. Tim Sweeney states this himself. I'll believe the engines developer before some nVidia PR blog.

nVidia couldn't have paid for anything else but a lockout in this case. No assistance would have been necessary from AMD to enable it. AA is a feature of the engine itself.

[steveroger]

Batman controversy is annoying. ATI, Nvidia and the game developer/publisher screwed the consumer. They all couldn't get their act together so that the game would work equally well in terms of access to AA? How do you come to this conclusion, just decide that everyone is lying or that everyone is telling the truth or some are lying and some tellng the truth--any way you look at it no one was looking out for the customer. And who is the customer? Any potential buyer of the game that has a high end gaming pc.

I am annoyed about Need for Speed Shift too. I was about to buy the game and found that it had some performance issues with ATI cards. It was all over the forums. I sent an email to EA support asking if the problems were true and whether I should buy the game now or wait for a patch. They answered: buy the game and wait for a patch. LOL.

What do I think of PhysX? I think it is mostly a publicity stunt designed to get me to buy Nvidia cards. But it is something that I would like to have anyway. LOL. I have an ATI card right now. If I bought a 8 seriers card could I put it in my PC and get the benefit of PhyX? I understand I can, but it isn't easy. If Nvidia wasn't so interested in squeezing every last dollar from me they would make it real easy for me to do this.

I suppose I can solve this problem just like I solve the same problem as it relates to the consoles. We have all the consoles and buy the best version of the game as tied to each console. Batman was best on the PS3 due to the playable Joker. So, I guess I will have to have two PCs on ATI and one Nvidia. Or can I run both cards in the same PC?:p

[bes]

I still got an average 50FPS with 12X edge detect AA with a HD4850 with the batman game for the AA thing didnt bother me very much..

I would like to see this card in action though...but I still wouldnt get one... I dont Like Nvidia ..even if the new Nvidia card blows the doors clean off the new ATI cards..

Since all games should run on the new ATI cards with maxed out settings for a long time anyway since my HD4850 still can with most games..

[steveroger]

I still got an average 50FPS with 12X edge detect AA with a HD4850 with the batman game for the AA thing didnt bother me very much..

I would like to see this card in action though...but I still wouldnt get one... I dont Like Nvidia ..even if the new Nvidia card blows the doors clean off the new ATI cards..

Since all games should run on the new ATI cards with maxed out settings for a long time anyway since my HD4850 still can with most games..

My point really is more along the lines is that there doesn't need to be a controversy at all and the card makers, developers and publishers should just get their act together.

[Baron_Fel]

Tamasi later went onto say that "no game developer on the planet is going to let us do anything to a game which prevents it from running on ATI, or having a good experience. Whenever we go to do something, the first principle we apply is 'do no harm' - you never make it worse than before you went in. Ever."

"If we did that, next time, the developer is going to say 'sorry, we don't want to work with you guys' and that's the end of our existence," added Rege.

This quote is a hell of a lot more believable than nvidia actively trying to sabatoge AMD. Like steveroger said, its more likely the blame lies with everyone on this one.


What do I think of PhysX? I think it is mostly a publicity stunt designed to get me to buy Nvidia cards. But it is something that I would like to have anyway. LOL. I have an ATI card right now. If I bought a 8 seriers card could I put it in my PC and get the benefit of PhyX? I understand I can, but it isn't easy. If Nvidia wasn't so interested in squeezing every last dollar from me they would make it real easy for me to do this.

Physx is GREAT for the...2 real games its implemented well in, which are Mirror's Edge and Arkham Asylum. It just isnt prolific enough to warrent a purchase of a nvidia card. There are far more worthwhile things to upgrade that do more for your overall experience, like a new monitor/speakers or more storage or an SSD.

Now, nvidia's stereoscopic 3D actually looks promising, but itll have to wait until 120hz monitors are cheap and high res before itll become relevant.

[Pickanewnameplzkthx]

Guys, it's real simple with ATi cards. They come pre equipped with driver issues, and they've been plagued with AA issues. Now I know the phanbois are gonna come out in force on this post, but that's the way it's been for a long time, and after the takeover by AMD it got worse. They cannot get their damn drivers fixed. I remember AA issues with those cards as far back as the X800's and that was years and years ago. I remember this stuff because I had an old All In Wonder card and got caught up in that fiasco, and then the driver problems started.

That old All in wonder card I had was the last ATi card I bought, because ATi hasn't come out with a performance leader since when I was ready to buy, and they can't get their damned drivers fixed. If they could do that, maybe I'd start to look at them again.

[Baron_Fel]

Guys, it's real simple with ATi cards. They come pre equipped with driver issues, and they've been plagued with AA issues. Now I know the phanbois are gonna come out in force on this post, but that's the way it's been for a long time, and after the takeover by AMD it got worse. They cannot get their damn drivers fixed. I remember AA issues with those cards as far back as the X800's and that was years and years ago. I remember this stuff because I had an old All In Wonder card and got caught up in that fiasco, and then the driver problems started.

That old All in wonder card I had was the last ATi card I bought, because ATi hasn't come out with a performance leader since when I was ready to buy, and they can't get their damned drivers fixed. If they could do that, maybe I'd start to look at them again.

This post is funny because the 4000 series generally does AA with a smaller performance hit than nvidia's offerings.

[renegadeafk]

Guys, it's real simple with ATi cards. They come pre equipped with driver issues, and they've been plagued with AA issues. Now I know the phanbois are gonna come out in force on this post, but that's the way it's been for a long time, and after the takeover by AMD it got worse. They cannot get their damn drivers fixed. I remember AA issues with those cards as far back as the X800's and that was years and years ago. I remember this stuff because I had an old All In Wonder card and got caught up in that fiasco, and then the driver problems started.

That old All in wonder card I had was the last ATi card I bought, because ATi hasn't come out with a performance leader since when I was ready to buy, and they can't get their damned drivers fixed. If they could do that, maybe I'd start to look at them again.
There is absolutely nothing wrong with atis drivers, I have used an x300, x1950pro, and now a 4870 and have never had one issue with their drivers. Some people have issues with both nvidia and atis drivers, thats just the way it is but they are on par with each other. I can't stand it when people like you spew this drivers are bad nonsense without even knowing what the hell they are talking about. I see this moronic nonsense posted all the time from both the nvidia and ati camp saying omg xxxx drivers suck becuase I had one random bad experience with them ten years ago!

You keep repeatedly mentioning phanboi's in your posts but what you don't seem to realize is that one of the biggest phanbois of all is you, I know as soon as you read this post you will probably call me a phanboi or some other equally immature term and try to dismiss anything I said, but your post is just pure hyperbole.

"because ATi hasn't come out with a performance leader since when I was ready to buy" What? ati and nvidia have been trading the performance crown over the years... they have both had "performance leaders"

And you seem to forget, right at this moment with the 5xxx series ati has quite the performance leader. (Although, yes technically nvidia does have the monster GTX 295 but it is a dual card solution)

And also overall price/performance is ALWAYS better than having the technically most powerful card. Whether its nvidia or ati I will always get the card that offers me the best performance for the price because unlike you, I am not a "phanboi"

I don't have any of these "aa" issues you speak of with my 4870 or any of my former ati cards. And I like how you pretend that nvidias drivers never have issues as well. Well have fun in your pretend world.

[Toto pectore]

T_Flight: I didn't have many problems with ATI drivers (1950GT, 3870 and now 4850) so I can't agree with you. Also things got a lot better with drivers after the takeover by AMD - especialy with drivers for Linux (but there is still lot of work).

[AlecJ32]

Great....More reason for me to try out '.NET Integration for CUDA'

[Pickanewnameplzkthx]

1

"because ATi hasn't come out with a performance leader since when I was ready to buy" What? ati and nvidia have been trading the performance crown over the years... they have both had "performance leaders"

And you seem to forget, right at this moment with the 5xxx series ati has quite the performance leader. (Although, yes technically nvidia does have the monster GTX 295 but it is a dual card solution)

And also overall price/performance is ALWAYS better than having the technically most powerful card. Whether its nvidia or ati I will always get the card that offers me the best performance for the price because unlike you, I am not a "phanboi"

I don't have any of these "aa" issues you speak of with my 4870 or any of my former ati cards. And I like how you pretend that nvidias drivers never have issues as well. Well have fun in your pretend world.

Ok, I won;t even respond to the driver issues. They are well documented, and not 3 posts above your is yet another ghame that just came out where Ati cards agauin have broken AA. Nvidia is not having those issues. That's a open and shut case.

Again, you've miseread my comments. I said "when I was ready to buy" that ATi has not had a pefromacne leader. I could care less about price to performance ratio. I'm not gonna sacrifice performance and longevity for price. That just means I have to buy more cards which makes it more expensive in the end.

A case in point is the card I have now. I could've bought one 300 dollar card, or 2x 250 dollar cards 6 months apart to be able to play the games I cureently am running at the frames I'm running them at. Let's see...500 dollars or 300 dollars for the same lecvel of performance. I'll take the 300 dollar single card please. Aw, and that has lasted me over a year. As for the 5870's? If they have better performance than Nvidia's card when I'm ready to buy again, they're bought. It's as simple as that. That pretty much quells your phanboi argument doesn't it?

Now, let's get into the hardware and the actual price to performance ratio. You get what you pay for. Want a 250 card? You'll get exactly 250 dollars worth of card. You'll get a IC board with less layers, you'll get lower rated caps, and you'll get VRM's that are rated for 90C instead of 130C. Doubt this? Go to XtremeSystems and check out the threads of those components that let the magic smoke out when they ran them on Furmark. Check out the recapping threads, and the guys looking to source new VRM's, and the retracing with wire due to burned out traces. Aw, and that's before they came out with a gimped driver that slowed the cards down when the Furmark execute was ran. They couldn;t handle the heat, and that particular program is what serious enthusisasts like myself use to test for stability of their OC's, and to make sure their cooling is up to par. I had no such issues with Nvidia. They were not overriding clock settings to make my attempts at checking for stability futile, and I had no overheating of VRM's. The 130 C rated VRM's never exceeded 80C on air, and with the Unisink I installed those temps went down to 60 C on 3 of the VRM's and the 4th never exceeded 65 on any test.

So again, if Ati gets their driver act together, and has the better card after Nvidia's release it's bought. You won't catch me making statements like "I don't care if _____ has the better card, I won't buy it." If the phanboi shoe fits wear it, but if you wanna come here calling me a phanboi you best come forth with something better than BS, and conjecture, because it don't apply.

[bes]

Ok, I won;t even respond to the driver issues. They are well documented, and not 3 posts above your is yet another ghame that just came out where Ati cards agauin have broken AA. Nvidia is not having those issues. That's a open and shut case.


Your an odd person ..still spewing bull crap..

Who gives a rats ♥♥♥ if I have to run the Batman game with my cards control panel AA ... im still getting maxed performance with 12X Edge Detect AA... I bet thats higher than the in game settings you would get with an Nvidia card..

Its not broken it works very well..

As for the rest of your post ... there has already been benchmarks showing the new ATI cards a performance leader over Nvidias best ...and AMDs next cards will be pretty uber also..

Nvidias new card with Fermi however probably will make them the leader for a bit ...they switch all the time..

The ONLY driver issues AMD/ATI have is with crossfire support ....I wont bother using crossfire because I have tried it with 3 different cards and I still didnt get any higher performance..

I also get ATI because they are cheaper normally..

[renegadeafk]

Ok, I won;t even respond to the driver issues. They are well documented, and not 3 posts above your is yet another ghame that just came out where Ati cards agauin have broken AA. Nvidia is not having those issues. That's a open and shut case.


What aa issues? you still provide no proof. I play TONS of games and all with 4xaa flawlessly, any "aa" issues are the fault of games devs like batman arkham asylum.

Again, you've miseread my comments. I said "when I was ready to buy" that ATi has not had a pefromacne leader. I could care less about price to performance ratio. I'm not gonna sacrifice performance and longevity for price. That just means I have to buy more cards which makes it more expensive in the end.

A case in point is the card I have now. I could've bought one 300 dollar card, or 2x 250 dollar cards 6 months apart to be able to play the games I cureently am running at the frames I'm running them at. Let's see...500 dollars or 300 dollars for the same lecvel of performance. I'll take the 300 dollar single card please. Aw, and that has lasted me over a year. As for the 5870's? If they have better performance than Nvidia's card when I'm ready to buy again, they're bought. It's as simple as that. That pretty much quells your phanboi argument doesn't it?

Now, let's get into the hardware and the actual price to performance ratio. You get what you pay for. Want a 250 card? You'll get exactly 250 dollars worth of card. You'll get a IC board with less layers, you'll get lower rated caps, and you'll get VRM's that are rated for 90C instead of 130C. Doubt this? Go to XtremeSystems and check out the threads of those components that let the magic smoke out when they ran them on Furmark. Check out the recapping threads, and the guys looking to source new VRM's, and the retracing with wire due to burned out traces. Aw, and that's before they came out with a gimped driver that slowed the cards down when the Furmark execute was ran. They couldn;t handle the heat, and that particular program is what serious enthusisasts like myself use to test for stability of their OC's, and to make sure their cooling is up to par. I had no such issues with Nvidia. They were not overriding clock settings to make my attempts at checking for stability futile, and I had no overheating of VRM's. The 130 C rated VRM's never exceeded 80C on air, and with the Unisink I installed those temps went down to 60 C on 3 of the VRM's and the 4th never exceeded 65 on any test.

I have no clue what you are talking about there it is completely incoherent, what are you trying to say? My ati card will just explode because it has "lower rated caps" this is pure fanboi garbage.

So again, if Ati gets their driver act together, and has the better card after Nvidia's release it's bought. You won't catch me making statements like "I don't care if _____ has the better card, I won't buy it." If the phanboi shoe fits wear it, but if you wanna come here calling me a phanboi you best come forth with something better than BS, and conjecture, because it don't apply.
You are just grasping at straws when you start pulling "nvidia has higher rated caps/vrms" out of your ♥♥♥

[Pickanewnameplzkthx]

Thanks for that remimder Bes. Ati will also have to deliver the XFire port eneable they've been promising for how long? I'm definiotely going with two cards, so no working link? No sale. Any other shills wanna come forward for ATi?

[renegadeafk]

SLI and crossfire scaling depends pretty much completely on the game, they can both scale perfectly well depending on the game.
http://www.overclock3d.net/reviews.php?/gpu_displays/xfx_gtx_260_xxx_sli_performance/5

You just continue to put out this self righteous nonsense. I am done arguing with a troll like you, anyone with a lick of common sense won't believe the crap you are saying anyway.

[Pickanewnameplzkthx]

SLI and crossfire scaling depends pretty much completely on the game, they can both scale perfectly well depending on the game.
http://www.overclock3d.net/reviews.php?/gpu_displays/xfx_gtx_260_xxx_sli_performance/5

You just continue to put out this self righteous nonsense. I am done arguing with a troll like you, anyone with a lick of common sense won't believe the crap you are saying anyway.

I still do not see where they enabled the XFire link in that review. Am I missing something? Could it be that the XFire link was another ad gimmick, and doesn't offer any measureable performance benefit, so they decided to hide it? Could be. If not, then surely they would want to eneable that right away. They keep promising it on every card, but so far have not delivered.

[renegadeafk]

I still do not see where they enabled the XFire link in that review. Am I missing something? Could it be that the XFire link was another ad gimmick, and doesn't offer any measureable performance benefit, so they decided to hide it? Could be. If not, then surely they would want to eneable that right away. They keep promising it on every card, but so far have not delivered.

Edit 2: that was the right link in my post, you had me thinking I copied the wrong one or something, again can you read?
http://www.overclock3d.net/reviews.php?/gpu_displays/xfx_gtx_260_xxx_sli_performance/5
This is one example from one of the benchmarks
4870
111.5 fps
4870 xfire
199.1 fps

GTX 260
113
GTX 260 SLI
163


If you can't read a benchmark there is no hope for you. As you can plainly see from 2 seconds of looking at it it tests single and both crossfire and sli. I think your fanboy blinders may have prevented you from looking at it at all.

[Pickanewnameplzkthx]

Can you even read? This is one example from one of the benchmarks
4870
111.5 fps
4870 cfire
199.1 fps

If you are too stupid to be able to read a benchmark you a truly hopeless.

Ok, I think we have a individual here that has a failure to read English...lemme speak slower...

W-h-e-r-e i-nnnnnn t-h-a-a-a--t r-eee-v--i-e-w -d-o-e-s i-t s-a-y t-h-a-a-a-t -t-h--e-e-e- C-r-o-sss-f-i-r-e l-i-nnn-k h-aaaa-s-s-s b-ee-e-n-n-n e-e-e-n-a-a-a-a-b-l-l-e-e-d?

W-h-a-a-a-a-t d-o-o-o-o t-h-o-o-o-o-o-s-e n-u-u-u-u-m-b-e-r-s h-a-a-a-v-e t-o-o-o-ooo d-o-o-o-o w--i-i-i-i--t-t-t--h-h-h -t-h-e- C-r-r-r-o-o-o-o--s-s-s-s-f-f-f-i-i-i-i--r-r-r-r-e-e-e-e l-ll-l-l-l-l-l--i-i-i-i-i-n-n-n-n-k-k-k-k?

We might need a radio. I'll contact NASA becasue we may have a case of selective hearing here.

[renegadeafk]

Its ridiculously easy to see, it shows crossfire fps and single card fps. Even a screenshot with pretty circles to make it easier for you. I don't get how you don't see this, it takes a true case of stupid because its staring you right in the face.
http://i33.tinypic.com/6rspci.png

We might need a radio. I'll contact NASA becasue we may have a case of selective hearing here.
Also your gradeschool insult makes no sense, what does nasa have to do with selective hearing or radios?

[rotNdude]

The type of back and forth crap being exhibited by the above two members is not what this thread is about. If we can't keep these threads civil and constructive, then I will raise the bar on what can and can't be posted in this section of the forum. I tend to be rather lenient, but that may be coming to a bone-shattering halt if this type of thing continues when product announcements or new information is made available.

Cheers!

[LordEC911]

GF100 is LATE. Nvidia even said so themselves.

The sources I was referring to earlier have some people they know in the industry.

G300 tapedout in June, they just got first silicon back at the end of August, give them a couple weeks for debugging and they will send it back to TSMC for a respin. Wait about 6-8weeks to receive the silicon back, we are looking at the middle of Nov. If all is well with that silicon they will ramp production and might get a handful of cards out before the end of the year.

Those same sources that are stating the less than 10% yields are also stating that G300 will have similar clockspeed problems like G200 due to power consumption.

Take it or leave it. It is solid information from a solid source. We will see who is right in a few months.

As far as Fudo being in Nvidia's pocket-
http://www.fudzilla.com/images/stories/2009/August/General%20News/fudoijensen.jpg

and all he is posting now is pro-Nvidia stuff.

[Baron_Fel]

GT200 has clockspeed problems? News to me.

[IcarusNine]

Their clock speeds were only a problem in the sense that they couldn't command the prices they wanted, and their yields didn't allow them to enjoy the prices they had.

Ok, I think we have a individual here that has a failure to read English...lemme speak slower...

W-h-e-r-e i-nnnnnn t-h-a-a-a--t r-eee-v--i-e-w -d-o-e-s i-t s-a-y t-h-a-a-a-t -t-h--e-e-e- C-r-o-sss-f-i-r-e l-i-nnn-k h-aaaa-s-s-s b-ee-e-n-n-n e-e-e-n-a-a-a-a-b-l-l-e-e-d?

W-h-a-a-a-a-t d-o-o-o-o t-h-o-o-o-o-o-s-e n-u-u-u-u-m-b-e-r-s h-a-a-a-v-e t-o-o-o-ooo d-o-o-o-o w--i-i-i-i--t-t-t--h-h-h -t-h-e- C-r-r-r-o-o-o-o--s-s-s-s-f-f-f-i-i-i-i--r-r-r-r-e-e-e-e l-ll-l-l-l-l-l--i-i-i-i-i-n-n-n-n-k-k-k-k?

We might need a radio. I'll contact NASA becasue we may have a case of selective hearing here.
Quoting this for posterity. It's unintentionally hilarious.

[LordEC911]

GT200 has clockspeed problems? News to me.
You really think they wanted to release a 933GFlop card?