nVidia : Annoucement Monday morning, 9am PST

[evolucion8]

(NV is competitive below $150, and these people don't know what a Direct X is).
NV can keep mindshare without too much problem when they have products which are still appealing, and they do for the majority of the market by volume.

Yeah, with their GTS 250.

Oh that's great, an announcement to another announcement, then in Pax will announce the availability of the cards and their performance numbers, the same that has been floating around the web for ages, then they will announce another date to announce the release date of the another announce which will have the release date of another announcement showing real cards instead of wood cards and the date of release of the announcement of Fermi's release date.

 

[Genx87]

I'm glad that it is Ati who is surviving the 3D graphics accelerator wars that started back in 95. Rendition, S3, Matrox, 3Dfx, and now nVidia. RIP nVidia.

A company who turns a healthy profit and controls 63% of the discrete graphics market is now RIP? A bit dramatic are we? This is like the 5800 fiasco except I actually expect this to perform better than the competition. If Nvidia survived that debacle. They will survive this one.

 

[MarcVenice]

I wasn't expecting this.

 

[pmv]

I'm glad that it is Ati who is surviving the 3D graphics accelerator wars that started back in 95. Rendition, S3, Matrox, 3Dfx, and now nVidia. RIP nVidia.

Why? I'd far rather there were at least two survivors. The particular survivor seems infinitely less important than the _number_ of survivors. All corporations are kind of interchangeable in the long run, they are ultimately just collections of mercenaries after all (much like sports teams).

 

[Qbah]

Once the cards become available, I think all the green folks will just totally ignore the 6 month delay and start yelling "It's here, grab it!". And all the people waiting 6 months to have a card 10% faster than the competition had for the past 6 months (my wild guess atm) will be very very upset. They could enjoy gaming for the past 6 months but instead listened to the hype.

I think the only thing the green team will expect is just to shrug it off, be happy that it's available and ignore all the hype, delays and let downs way that happened along the. Wouldn't be the first time (that goes for both teams)...

 

[aigomorla]

what is wrong with 5970 drivers? I heard that they the new cats had worked out most of the kinks.

Lets wait until 10.3 to see.
I heard its still fubarded compared to a single HD5870.

This is what my friends tell me whose been benching these cards.
Great bench cards, but the 5870's i heard are better in raw gaming.

How so? Well, by the time I can actually get a 5970 for less than $650 they will probably have good drivers out. Unfortunately I cannot do crossfire 5870's.

Thats what im crossing my fingers for... by then fermi should be out...
So a fully matured 5970 vs. a hidden fermi is the options im left with.

Thats nothing other than a gigantic piss take.
Seems i'll be waiting longer than i thought to replace my 4890.

Well im waiting to swap my 2 x HD4870X2's... and as u can see, there not that much slower then the competition right now.

Raw decision factor is... do i want eyefinity, (i have the monitor setup for it), or do i want to give nvidia a shot again....

 

[SlowSpyder]

Once the cards become available, I think all the green folks will just totally ignore the 6 month delay and start yelling "It's here, grab it!". And all the people waiting 6 months to have a card 10% faster than the competition had for the past 6 months (my wild guess atm) will be very very upset. They could enjoy gaming for the past 6 months but instead listened to the hype.

I think the only thing the green team will expect is just to shrug it off, be happy that it's available and ignore all the hype, delays and let downs way that happened along the. Wouldn't be the first time (that goes for both teams)...

Yup, we've been hearing how we should wait for Fermi pretty much since the launch of the 58xx cards. It looks like if you didn't wait and went with the 58xx, you would have been enjoying gaming on your new card for up to 5 months now as opposed to waiting for announcements of announcements.

 

[Rezist]

So what was the annoucement?

 

[Apocalypse23]

Down with Nvidia!

 

[NoQuarter]

So what was the annoucement?

That they'd have a booth at PAX East.

 

[pmv]

So what was the annoucement?

As far as I can make out it was "at a future date there will be another announcement".

Its all very well to joke about people waiting for fermi because they are nvidia fans, but some of us don't care which company has what, we are simply waiting for prices to come down. I suspect that will be an even longer wait than waiting for Fermi (they really should have codenamed the chip 'Godot')

Who do people feel more sorry for? Nvidia's engineers, who must haven been under tremendous pressure (and, notwithstanding recent difficulties are obviously very smart guys who I for one would be fascinated to hear from), or nvidia's marketing department who must also be working overtime desperately trying to spin a story out of almost nothing at all?

 

[SHAQ]

They are having a raffle for a PC with 480 SLI at PAX. Will the winner be able to get the PC that weekend or do they have to wait for launch? lol

 

[Hauk]

Let this be known as nVidia's most foolish day..

 

[PingviN]

Let this be known as nVidia's most foolish day..

22/2 - Nvidia fools day

Free for anyone to act attention-whore in memory of this great announcement!

 

[MagickMan]

So, NV had nothing new and Charlie was talking out of his ass (again)? Sounds like a regular Monday, to me.

I'm through with giving the Green team more time, they can piss off. As of right now I'm officially shopping for a 5870

 

[Zargon]

That they'd have a booth at PAX East.

ROFL. so simple its hilariously amusing

 

[scooterlibby]

Maybe I'll get a C2Q to replace my e8400 rather than a 5970 and wait/save for as long as its takes for the 5 series refresh.

 

[SHAQ]

This fiasco has had a nice effect on the 295 used prices on Ebay. Yesterday one went for $580 with shipping. I guess I'll sell mine and get a profit and go back to 260 SLI until the 480 comes out.

 

[cbn]

A company who turns a healthy profit and controls 63% of the discrete graphics market is now RIP? A bit dramatic are we? This is like the 5800 fiasco except I actually expect this to perform better than the competition. If Nvidia survived that debacle. They will survive this one.

Even if Fermi is hotter than Nvidia would like the architecture itself may end up being very long lasting.

The fact even Charlie admits the tessellation performance is outstanding makes me hopeful that future interations of this chip will deliver where we want it to deliver most--->3D.

I myself would rather give up peak frame speeds if it meant the dips (ie, minimum frame rates) were substantially better. I'm sure Nvidia's 3D vision works better with this chip for that reason also. Doesn't 25 FPS look like 12.5 FPS with the Nvidia glasses on?

 

[Genx87]

Even if Fermi is hotter than Nvidia would like the architecture itself may end up being very long lasting.

The fact even Charlie admits the tessellation performance is outstanding makes me hopeful that future interations of this chip will deliver where we want it to deliver most--->3D.

I myself would rather give up peak frame speeds if it meant the dips (ie, minimum frame rates) were substantially better. I'm sure Nvidia's 3D vision works better with this chip for that reason also. Doesn't 25 FPS look like 12.5 FPS with the Nvidia glasses on?

As bad as the NV30 was the NV35 wasnt "that" bad. But the work on those chips setup the NV40 that set Nvidia rolling until this debacle. They got bit by the process bug again. A little aggressive. It will be interesting to see how this unfolds in 2 years.

 

[cbn]

As bad as the NV30 was the NV35 wasnt "that" bad. But the work on those chips setup the NV40 that set Nvidia rolling until this debacle. They got bit by the process bug again. A little aggressive.

What do you mean by this? Are you talking about Nvidia's large die strategy.

It will be interesting to see how this unfolds in 2 years.

Yes I think it will be interesting to see how the next 2 years unfolds. Will ATI press ahead with larger and larger dies hoping to double performance every generation? Or will they change architecture in order to gain more performance out of less silicon?

 

[Genx87]

What do you mean by this? Are you talking about Nvidia's large die strategy.

NV30 was a bold design. The work and pains of that chip provided years of prosperity for Nvidia. What I mean by they got bit by the process bug was Nvidia is bringing out a bold arch this generation that required the newest process technology to hit their targets. Just like the NV30. And just like the NV30 their manufacturer failed. Biting Nvidia in the ass. It will be interesting to see how Nvidia copes with this set back. They were pretty damn quick with the NV35 which helped until the NV40 a year later opened up the flood gates.

This is so much like the NV30 it is like deja vu at this point.

Yes I think it will be interesting to see how the next 2 years unfolds. Will ATI press ahead with larger and larger dies hoping to double performance every generation? Or will they change architecture in order to gain more performance out of less silicon?

I think it depends on where gaming goes and if ATI wants to get involved in the HPC market. If tesselation takes off in game engines they are forced to spend transistor counts in that dept. They may even have to go to a MIMD arch like Nvidia to get better performance. And of course they will need to up their double precision performance for the HPC market. And have they offered a better anistrophic filtering technique? They did that aggressive angle dependent AF in the R300 and havent looked back.

 

[dguy6789]

Will ATI press ahead with larger and larger dies hoping to double performance every generation?

Absolutely not. ATI has already learned that that is not the way to go with the R600. Intel learned it with Prescott. Nvidia is learning it right now(one would think they learned from NV30 though). Small dies are usually very energy efficient, cheap to manufacture, produce good yields, and you can just use two of them for the ultra high end. That's the way to go. Expect ATI to do that for the foreseeable future.

When you go with a large die strategy, costs rise astronomically. They are far more difficult to manufacture, yield problems are much worse, power consumption rises, it's all quite troublesome.

This page is pretty informative on some of the problems with using a large die.

http://www.anandtech.com/video/showdoc.aspx?i=3740&p=8

I fully expect Nvidia to implement their own small die strategy in the future.

 

[zerocool84]

Makes me even more happy I didn't wait. I've been a die-hard Nvidia fan for a long time but got sick of the waiting and lack of info. I'm enjoying awesome gaming now

 

[Genx87]

Absolutely not. ATI has already learned that that is not the way to go with the R600. Intel learned it with Prescott. Nvidia is learning it right now(one would think they learned from NV30 though). Small dies, very energy efficient, cheap to manufacture, just use two of them for the ultra high end. That's the way to go. Expect ATI to do that for the foreseeable future. I fully expect Nvidia to implement their own small die strategy in the future too.

I guess it depends on the level of process technology and what one wants to do with a chip. If for instance you need to have two chips to do the work of one. You also have to build an interconnect that can do such work. That adds costs into the R&D and PCB manufacturing. Nvidia with their desire to get into HPC markets probably isnt going to be that interested in scaling down their chips and doubling up. It makes more sense to have a large single die that can it all.

We like to think this hiccup is the norm but it isnt. Nvidia has been burned twice in a decade. The other 8 years they have steamrolled ahead with the bigger single die approach.