NV30 not taped out yet....

[RussSchultz]

Re: tapeout to production

Here are the tapeout to production time periods for the past Nvidia chips:

Riva TNT2 - 98 days
GeForce 256 - 104 days
GeForce2/GTS - 102 days
GeForce/MX - 110 days
GeForce3 - 118 days
XGPU - 120 days
GeForce4 - 101 days

So best case if they tape out this week then they will be in production by mid-November.

I wonder what those dates really represent. Initial tapeout to start of production? Design finalization to production wafers out of the fab? Or any combination of the above 4 items.

 

[KimB]

A "tape out" is where you download your physical design databases so that you can create the masks needed for production. Generally, the completion of a tape out is the creation of the masks themselves, which generally runs several hundred thousands of dollars.

But was he talking about final production masks? Or first-attempt masks? (whatever you call it...).

 

[RussSchultz]

Gunhead, Geek apparrently took what I wrote as some sort of assertation that there's no problems in the NVIDIA world and jumped all over me. John edited his rant out, and put his name there to show he edited it, but it looks kinda funny...like John is being addressed and not the other way around.

Anyways to address the original confusion, perhaps I didn't add enough animated smiley faces and punctuation to fully emote perplexity and whimsy in my "maybe its a sick disinformation campaign" comment.

The amount of rumors and "facts", confirmations and anti-confirmations, speculations and wild assertations swirling around the NV30 is staggering and perplexing. One day we have one story, the next another.

To be clear: I make absolutely no assertations whether NV30 will or will not be late, especially with as little concrete information that we have. I really don't care either way.

 

[Guest]

This info could fit with the "move to UMC because TSMC sux0rs" rumor, or maybe not.

Such very strange rumors(has taped out, hasn't, .15u, .13u, TSMC, UMC), with confirmations, opposing rumors and opposing confirmations.

I wonder if the real story is a design cycle gone horribly wrong, or some brilliant, yet sick, joke of a marketting disinformation campaign.

Anyways, if they're going to make it for for the holiday season (which means product on the shelves, in my company), they must have wafers entering production right now or within a few weeks at the most. TSMC is supposedly relatively full and has close to a 2-3 month backlogs (though NVIDIA is a big company, so maybe not for them), so much later won't give their OEMs time to manufacture and ship them in time to get them to the shelves.

Of course, Chinese New Year is also a holiday season...

No Russ it isn't a "sick joke" I just heard the CEO of nvidia say the NV30 has not taped out. This is not a game, nvidia stock just lost massive valuation.

The only "marketing disinformation" is that the NV30 has taped out rumor that we have been hearing for the last while. Further the Pictures that were floating around were totally bogus along with some other missinformation spewing from nvidia regarding its specs found in the fact that the NV30 can only handle "256" static instructions not "1024".

Further I doubt very much you will have this card for Christmas Russ. You may get a paper launch, then another month or so later be able to buy it in January/Febuary depending on the demand I supose. By that time ATI surely would have a refresh of the Radeon 9700.

The problem is with their design cycle and ATIs ability to put out a powerful DX 9 card at the .15um process where nvidia did not have this capasity and was forced to go with their NV30 design @ the .13um process in a rather rushed manner at that. This does not mention the fact that the .13um process still has issues.. Their real problem is ATI.

EDIT: This is basicly the same "rant" that was edited out with a few extra points and less colorful names for Russ. I hope I don't offend anyone with it.

 

[DemoCoder]

NVidia's valuation is based on their mainstream chipsets, the xbox, and the workstation market, not their high end chipsets. Why a late NV30 spells doom for them, I don't get. If you look at their balance sheet, even the GeForce3/4 isn't their main revenue generator.

The only thing that will really hurt NVDA's fundamentals is killing the GF2MX/GF4MX and continually declining X-Box sales.

 

[3dcgi]

Where did those tapeout numbers come from? If 120 days is their worst case from initial tapeout I'm very impressed. If the nv30 is just getting ready to complete initial tapeout then nvidia has 0 margin for error. The slightest slip up and they won't ship in time for Christmas.

 

[Guest]

NVidia's valuation is based on their mainstream chipsets, the xbox, and the workstation market, not their high end chipsets. Why a late NV30 spells doom for them, I don't get. If you look at their balance sheet, even the GeForce3/4 isn't their main revenue generator.

The only thing that will really hurt NVDA's fundamentals is killing the GF2MX/GF4MX and continually declining X-Box sales.

They are going to lose the design wins that come with the company that holds the high end market. Nvidia should start to lose market share. I am not saying this is the end of nvidia. Granted it may be the end of their glory days though. The Radeon 9000/Pro are poised to take the market that the gf2mx/gf4mx occupy. The top end graphics card(Meaning the company that carries the performance crown for a prolonged time periode like we are about to see ATI do.) maker garners much attention in the OEM market this is how nvidia has managed such a large market share for some time. Now as a result of the loss of the high end market their low end market is in serious jeaprody considering how poor their low end products really are. How much will nvidia loose and how much will ATI gain is the question to ask at this point IMHO.

 

[X-Reaper]

If 120 days is their worst case from initial tapeout I'm very impressed. If the nv30 is just getting ready to complete initial tapeout then nvidia has 0 margin for error. The slightest slip up and they won't ship in time for Christmas.

Thats on proven technology. .13u is new for Nvidia. I doubt Nvidia could tape it out that soon. But we shall see.

 

[Guest]

If 120 days is their worst case from initial tapeout I'm very impressed. If the nv30 is just getting ready to complete initial tapeout then nvidia has 0 margin for error. The slightest slip up and they won't ship in time for Christmas.

Thats on proven technology. .13u is new for Nvidia. I doubt Nvidia could tape it out that soon. But we shall see.

I would tend to agree. It makes sence that a more complex process would cause them to take longer. But I am sure we will hear all sorts of other explainations as to why it won't for some reason or other.

 

[RussSchultz]

If 120 days is their worst case from initial tapeout I'm very impressed. If the nv30 is just getting ready to complete initial tapeout then nvidia has 0 margin for error. The slightest slip up and they won't ship in time for Christmas.

Thats on proven technology. .13u is new for Nvidia. I doubt Nvidia could tape it out that soon. But we shall see.

I think some of those numbers were for products on similarly unproven technologies. (3dfx went .25u for VSA100, for example, while GF was in .18u (or was it .22u?)) Much doom and gloom was predicted by 3dfx cheerleaders, but never happened.

Of course, doesn't mean they'll get away with it every time.

 

[DemoCoder]

The top end graphics card(Meaning the company that carries the performance crown for a prolonged time periode like we are about to see ATI do.) maker garners much attention in the OEM market this is how nvidia has managed such a large market share for some time. Now as a result of the loss of the high end market their low end market is in serious jeaprody considering how poor their low end products really are. How much will nvidia loose and how much will ATI gain is the question to ask at this point IMHO.

The only problem with your theory is that NVidia hasn't lost the performance crown yet (no R300 currently on shelves) and that even if it does lose the performance crown (I think we can safely predict this), it will only be for a few months (assuming NV30 performs as well as R300). NVidia has had the performance crown for years and they have had delayed cycles before.


I personally don't think what attracted OEMs to NVidia was the performance of their high end parts. I think what attracted OEMs was NVidia's consistent execution, timing new chips with new features to OEM release schedules. After all, the consumers technically don't buy "NVidia" GF2MX cards, but buy value cards with totally unrelated name and branding. I doubt the average GF2MX owner even knows what a GF4-4600 is!

So, NVidia could lose some design wins, not based on chipset performance, but merely because they miss an OEM cycle.

My personal opinion is that the R300 and NV30 are going to be so close in performance that it doesn't really matter performance wise. NVidia has missed Christmas before and still recovered.


I just don't see why so many people keep making this out to be some kind of war where one company has to win absolutely. The fact that we will have 2 DirectX9 parts soon with roughly equal performance and features is a HUGE BOON to us gamers. This is quickly becoming a 2-horse market like with Intel and AMD and do you honestly think we would be better off if either Intel or AMD was destroyed?

So why does it seem so many ATI fans are hoping for Nvidia to lose big time?

 

[nooneyouknow]

I can't believe people and their theories. Tape out commonly refers to the first silicon of the chip. Why would Nvidia change that? In fact, they can't. The reason why this issue was addressed is that they are legally required to address rumours that can impact their stock price, positively or negatively.

Now, they can still make X-Mas, but in reality, only if their 1st silicon is perfect. Highly doubtful considering the size of the chip. But, not impossible.

As far as OEM', etc.. They have a spring and a fall refresh typically. ATI hit the Fall, Nv30 will be good timing for Spring. That is the reality.

 

[jjayb]

The way he tried to tap-dance around the fact that it hadn't taped out yet, I would say it hasn't done an initial tape out. If it had already done some kind of initial tape out, I'm sure he would have gladly mentioned that. He tried not to have to say it didn't tape out, but the caller was persistent on getting an outright answer.

 

[jb]

Dont forget tape out != working parts on PCB boards. That could be even another up hill battle. And this is assuming that everything comes back OK from the fab. What do you think the typical run time is for some things big (110+ million transistors on .13 micron) 8 weeks?

 

[Ichneumon]

Dont forget tape out != working parts on PCB boards. That could be even another up hill battle. And this is assuming that everything comes back OK from the fab. What do you think the typical run time is for some things big (110+ million transistors on .13 micron) 8 weeks?

well... Nvidia's much touted "100 days" from tape-out to production would suggest at least 14 weeks. With that probably being a very optimisitic outlook for such a complex chip IMO.

 

[Pete]

I have to say, given the rumor and tip build-up the past few days, I wasn't expecting this.

ATi must be dancing a jig.

 

[echo]

The CFO (Christine Holmberg) stepped down.
R. Huddy and gang leaves the company to form codemafia
Nvidia starts hyping NV30 with paper specs (nvidiots start to debate about its performance, kind of silly really); a sign of desparation perhaps.
Nvidia says it will miss earnings, and that nv30 has not taped out.

Something's amiss?

 

[Hellbinder]

Yes...

 

[mpegleg]

Those numbers I gave for tapeout-to-production are almost certainly from final metal spin to production. The table I got them out of actually had one more column in it for mask rev. I'll post it tomorrow.

BTW I tend to agree there probably hasn't been an initial tapeout or else Jen-Hsun would have touted it (unless the silicon came back completely dead in which case its better to pretend there was no tapeout ).

 

[Nappe1]

This is quite an issue, isn´t it.

...Even worst for nvidia fans.

well, maybe they will get an idea how to feel being a Bitboys fan. (having at least quite few tape outs and when the release time comes, always something goes wrong...)

and yes, that was a joke.

back to subject, it looks like Big Train has some troubles to stay on track... and not so small ones.