Nvidia confirms the GTX 580

[Sam_oslo]

Because its a GTX580??? Maybe?
If thats all that a GTX580 is, it shouldnt name GTX485 or GTX490!
Dont get me wrong, I AM an nVIDIA fan, but I dont have my eyes shut...

The point is, that graph is not talking about performance. It's talking about noise.

Maybe it's a "fake" new generation, and can't fill the expectations we all have from a new generation. I've personally talked about this before, but we don't know it for sure yet, the real and final performance-gain remains to be seen.

But AMD has started the fake- generation-game with 68xx in this round, and maybe they will both try this number-game until they run out of numbers. We have to wait and see.

[Picao84]

The point is, that graph is not talking about performance. It's talking about noise.

Maybe it's a "fake" new generation, and can't fill the expectations we all have from a new generation. I've personally talked about this before, but we don't know it for sure yet, the real and final performance-gain remains to be seen.

But AMD has started the fake- generation-game with 68xx in this round, and maybe they will both try this number-game until they run out of numbers. We have to wait and see.

No, AMD released a NEW CHIP.. That they called it 6800 instead of 6700 has nothing to do with it. 6800 does have, in fact, better tesselation than 5800, among other features. This GF110 just looks like a GF100 done right, and with a bifier cooler.

[Sam_oslo]

No, AMD released a NEW CHIP.. That they called it 6800 instead of 6700 has nothing to do with it. 6800 does have, in fact, better tesselation than 5800, among other features. This GF110 just looks like a GF100 done right, and with a bifier cooler.

Nobody is saying 68xx didn't introduce changes, it has evolved and got better, but we all have expectations about performance-gain and power usage for a new generation, that usually requires a shrink to achieve.
As a minimum 6870 should beat 5870 to deserve it. It didn't happen, so it's a "fake" new generation, for sure.

It remains to be seen if Cayman or GTX580 can deserve a new first-digit, but AMD has already started the number-game in this round. So it won't surprise me if none of them could stand to the expectations for a new generation. We have to wait and see.

[Picao84]

Nobody is saying 68xx didn't introduce changes, it has evolved and got better, but we all have expectations about performance-gain and power usage for a new generation, that usually requires a shrink to achieve.
As a minimum 6870 should beat 5870 to deserve it. It didn't happen, so it's a "fake" new generation, for sure.

It remains to be seen if Cayman or GTX580 can deserve a new first-digit, but the number-game is on, and it won't surprise me of none of them could stand to the expectations for a new generation. We have to wait and see.

You would be right, IF 6800 was the higher performing single GPU. Which is NOT. The 6970 will be. Following that logic, there would have to be also a GTX590 SINGLE GPU, for things to be equal, and nVIDIA not be criticized. But, as far as we know, there ISNT.

[saaya]

noise and temps are very important to me, so i love that they focus on that...
but... i dont like how they present it as their own tech and invention and idear...
i mean vapor chambers have been around for quite a while...
am i the only one who thinks its weird that the only thing they talk about when it comes to the 580 is the heatsink? a third party component? hmmm

then again, its just a pre-launch leak...
which makes it a pre-pre-paper release
and a pre-pre-pre-release
and a pre-pre-prep-pre cards are available in some shops in some regions

hah... really funny huh?
in the 21st century both hw and software companies have managed to fully streamline introductions of new products
everything is available or presented a LONG time (to some extend) before retail availability

[Picao84]

What I cant understand is WHY they didnt go with this solution for original GTX480! That is the only thing that still gives me some hope. There must something else dammit
Or then, they are only admitting their own stupidity, which would be also a first

[highoctane]

What I cant understand is WHY they didnt go with this solution for original GTX480! That is the only thing that still gives me some hope. There must something else dammit
Or then, they are only admitting their own stupidity, which would be also a first

Well they achieved what they needed to with the 480 which was simply enough to take the single gpu pole position.

[Lanek]

What I cant understand is WHY they didnt go with this solution for original GTX480! That is the only thing that still gives me some hope. There must something else dammit
Or then, they are only admitting their own stupidity, which would be also a first

Just cause they have no more time to fix it, they was allready 5-6month late... at the moment they have find a way to release the GTX480 and 470, they have do it .... ( and it look like the solution, was to disable what was buggy ). then It seems they have continue to work on it and there you get the GTX580...

[zerazax]

What I cant understand is WHY they didnt go with this solution for original GTX480! That is the only thing that still gives me some hope. There must something else dammit
Or then, they are only admitting their own stupidity, which would be also a first

Or maybe they just messed up on the design of Fermi and it was really hard to get the process down. Now they've had some more experience, and the 580 can actually be produced at reasonable yields etc. but they needed a product out 6 months ago

[cowie]

This was in the rumor mill last week.
I think nv always wanted all the added functions and epp for the fermi.
----
Colleagues over at Nordic Hardware heard NVIDIA sacrificed the high-performance computing (HPC) functionality of Fermi to make the GeForce GTX 580 a smaller, more efficient card that's more suitable for the retail market. If the rumors are right, the card uses a 40nm GF110 GPU with 512 CUDA cores but has 300 million transistors less than the GeForce GTX 480 due to the canning of the HPC part. GPU clockspeed is said to be 775MHz, with 1544MHz shaders and 1536MB GDDR5 clocked at 4008MHz.

"According to sources it can be as many as 300 million transistors that NVIDIA has been able to cut in this way. The effect is that GF110 will be a GPU targetting only retail and will not be as efficient for GPGPU applications as the older siblings of the Fermi Tesla family. Something few users will care about.

The latest information that has reached us are said to be the final specifications for NVIDIA GeForce GTX 580. These were published at it168.com and strengthen the information we received earlier. There we can clearly see how NVIDIA beside processing units also raised the clock frequencies for both memory and GPU. This was made possible through a more mature circuit manufacturing where NVIDIA has managed to lower power consumption of GF100 with 10-15% without any physical alterations. This gave the comany more margin to work with."

[Pantsu]

I doubt there will be any availability problems due to low demand in this price category and the maturity of the process node. And of course a lot of people will simply wait for Cayman before making a decision. This still won't be the fastest card around when Nvidia touts it as the fastest GPU. 15-20% is easy with the full 512 cores and some architectural tweaks, no need to go about cutting any GPGPU stuff. I'd be shocked if something like that happened after Nvidia has put so much effort into Cuda marketing.

[Metroid]

It looks like is taking orders to deliver Monday

[Picao84]

I'd be shocked if something like that happened after Nvidia has put so much effort into Cuda marketing.

CUDA != Dual Precision, ECC, Large Caches, etc...
CUDA started with G80 and this chip didnt feature any of the above..

[Svnth]

"The water boils and it cools the fins." This guy has no idea what he talks about, a true marketing guru no doubt. I especially like the "a couple of years ago we had a part called the 285" well that was last year, he doesn't even know their own products.

But seems like a good cooler well just 2 more days.

Pretty fail post.. it was January of 09.. one year and 10 months ago. That's basically a couple of years.

Nice nitpick attempt though.

[Pantsu]

CUDA != Dual Precision, ECC, Large Caches, etc...
CUDA started with G80 and this chip didnt feature any of the above..

Of course I didn't mean dropping Cuda or something like that. Simply that touching the GPGPU performance would look bad when they have touted it so much.

[Olivon]

That will be quite interesting changing GF110 heatsink with GF100's one in order to see the efficiency of vapor chamber, no ?

[zerazax]

It looks like is taking orders to deliver Monday

Advertising 30% faster in Vantage and seeing the PR slides stating a quieter and cooler 480 does not bode well for in game performance as much as I was looking at...

How much is that price in comparison to the GTX 480 etc. so we can get an idea of how much this is going to cost in USD?

[cold2010]

[Chickenfeed]

Impressive if stable on air. Do like.

[RaZz!]

seems like someone really had some fun with the red paint brush, eh?

i mean... even the fanspeed? CMON... :<

[Flying duck]

I wanted to post this just so it stays public before Charlie re-edits yet another article.

"We've contacted multiple NVIDIA board partners about the upcoming GTX580, and it's looking like a storm..."

"Unfortunately for some of the manufacturing partners, it seems at this early stage that NVIDIA has failed to make them aware of it!

We contacted four of NVIDIA's top partners on the phone this afternoon, and we've paraphrased the conversations below:

Atomic: Hi {A}, we were wondering when we'd be able to get a review sample of the new GTX580? We were told to expect them roughly by the 9th of November.

Partner A: We're getting some in mid November. You'll be getting one then.

Atomic: Hi {B}, when will we be getting one of your GTX580 cards?

Partner B: We're not actually sure. Stock might be in by December, but we haven't got a confirmation at this stage.

Atomic: Hi {C}, when can we expect the new GTX580 card?

Partner C: ....We don't have that card on our roadmap. Does it exist?

Atomic: Hi {D}. So, how's about that new GTX580?

Partner D: We're getting stock in soon, so expect the card shortly."

http://www.atomicmpc.com.au/News/237...ch-flawed.aspx

[Sam_oslo]

"We've contacted multiple NVIDIA board partners about the upcoming GTX580, and it's looking like a storm..."

"Unfortunately for some of the manufacturing partners, it seems at this early stage that NVIDIA has failed to make them aware of it!

We contacted four of NVIDIA's top partners on the phone this afternoon, and we've paraphrased the conversations below:

Atomic: Hi {A}, we were wondering when we'd be able to get a review sample of the new GTX580? We were told to expect them roughly by the 9th of November.

Partner A: We're getting some in mid November. You'll be getting one then.

Atomic: Hi {B}, when will we be getting one of your GTX580 cards?

Partner B: We're not actually sure. Stock might be in by December, but we haven't got a confirmation at this stage.

Atomic: Hi {C}, when can we expect the new GTX580 card?

Partner C: ....We don't have that card on our roadmap. Does it exist?

Atomic: Hi {D}. So, how's about that new GTX580?

Partner D: We're getting stock in soon, so expect the card shortly."

http://www.atomicmpc.com.au/News/237...ch-flawed.aspx

LOL, good one.

I've been speculating that nVidia has been hiding this card (keeping it secret) to drop it from clear sky on Cayman.

Judging from these conversations, apparently they did a really good job, so good that even partners had no idea about the storm that is blowing on Cayman now.

[Picao84]

images

Cant understand what they are trying to show there

[Jowy Atreides]

"We've contacted multiple NVIDIA board partners about the upcoming GTX580, and it's looking like a storm..."

"Unfortunately for some of the manufacturing partners, it seems at this early stage that NVIDIA has failed to make them aware of it!

Actually, this seems correct when you look at the post above.

"Home delivery only. Direct from our main warehouse." etc.

Seems like nvidia HAS been shunning partners in favour of it's own brand supply chain

[Olivon]

Seems like nvidia HAS been shunning partners in favour of it's own brand supply chain

Really ?