NVIDIA GF100 & Friends speculation

[tannat]

No, you were actually talking about one specific product - the GTX 460. You're only changing your tune now after folks pointed out the flaws in your analysis. Of course operating profit is dependent on the ongoing success of the overall product line but that's not what you started out saying.

My point was really that there is no proof to either or neither, although i took a very pointy position.

It's also true that I took the GTX460 as example in one post, but only as example and that is quite obvious.

Are you seriously trying to tag me with something I did not express explicitly? You'r assuming some agenda here to make that kind of reasoning and I can't stop you. But it is nonsense nevertheless. And quite bad form btw.

 

[Arun]

trinibwoy: It's certainly not what he started out by saying, but I think he made it pretty clear he changed his position now that he understands the distinctions better, so you can hardly blame him for that I think

My point was really that there is no proof to either or neither, although i took a very pointy position.

It's also true that I took the GTX460 as example in one post, but only as example and that is quite obvious.

Are you seriously trying to tag me with something I did not express explicitly? You'r assuming some agenda here to make that kind of reasoning and I can't stop you. But it is nonsense nevertheless. And quite bad form btw.

I think everyone decided to pick on you specifically because you made the most clear-cut statement subject - and apparently that was mostly due to a misunderstanding between gross and operating margins. But now that you've reduced your position to a much more defensible one, you're still the one being picked upon by everyone disagreeing with this position even if there are quite a few others who would say the same.

So don't take it personally, at this point I don't think most people want to attack you personally - it's just a matter of convenience to target you, which I realise isn't the most polite option maybe but this is the Internet so I think it's about par for the course heh

Anyway: of course if you tried to dissociate the incremental R&D of various products, it is possible that some have not had/will not have sufficient gross profits to pay for their R&D. However, IMO given the likely gross margins involved throughout the product line, the probability that this is true for any current NVIDIA product is quite low. Keep in mind the incremental R&D for, say, GF106 is small compared to the total R&D required for GF1xx in general - a lot of it is shared.

Now if you wanted to 'share' the total architecture R&D between the various chips you very slightly increase the probability that one product wasn't worth its R&D, but that's also a completely useless metric because it implies there was an option in the first place not to work on a new architecture at all. Of course that was not an option, so only the incremental R&D for any given chip should be considered if you want to estimate its return on investment. And even that is very hard to estimate even with all the numbers (which we don't have) and quite subjective.

 

[tannat]

trinibwoy: It's certainly not what he started out by saying, but I think he made it pretty clear he changed his position now that he understands the distinctions better, so you can hardly blame him for that I think

I think everyone decided to pick on you specifically because you made the most clear-cut statement subject - and apparently that was mostly due to a misunderstanding between gross and operating margins. But now that you've reduced your position to a much more defensible one, you're still the one being picked upon by everyone disagreeing with this position even if there are quite a few others who would say the same.

So don't take it personally, at this point I don't think most people want to attack you personally - it's just a matter of convenience to target you, which I realise isn't the most polite option maybe but this is the Internet so I think it's about par for the course heh

Anyway: of course if you tried to dissociate the incremental R&D of various products, it is possible that some have not had/will not have sufficient gross profits to pay for their R&D. However, IMO given the likely gross margins involved throughout the product line, the probability that this is true for any current NVIDIA product is quite low. Keep in mind the incremental R&D for, say, GF106 is small compared to the total R&D required for GF1xx in general - a lot of it is shared.

Now if you wanted to 'share' the total architecture R&D between the various chips you very slightly increase the probability that one product wasn't worth its R&D, but that's also a completely useless metric because it implies there was an option in the first place not to work on a new architecture at all. Of course that was not an option, so only the incremental R&D for any given chip should be considered if you want to estimate its return on investment. And even that is very hard to estimate even with all the numbers (which we don't have) and quite subjective.

Thanks,

and this very much what I try to state; we don't know the profit ratios between cards and lines and we don't even now know however gross margins have been compromised (thanks for the very clear explanation previously btw). Decisions are long term and market segments are adjustable. I also made it clear that if (forever what reason) nvidia have chosen to compromise profit for market share, I (personally), don't think that is a big deal. On the contrary - such a step (if happened ) can have been very wise, although seemingly defensive short term

 

[tannat]

And Arun, I was obviously caught in a fire and have to bite the bullet

I was very pointy with the starting assumption, mainly since I saw the discrepancy in logic previously.

 

[trinibwoy]

Are you seriously trying to tag me with something I did not express explicitly?

I'm not picking on you but I thought you were pretty explicit.

It's enough to see that there are not really room for all the products to be profitable on their own.

That's a pretty clear statement that you believe some products aren't profitable. Though you contradicted that a few posts later with:

1) I've never claimed that Nvidia is selling products below costs.

I agree that there isn't enough information to make a definitive claim either way. Though I'm baffled why you think the evidence (45% margins) is in any way supportive of your notion that some products may be selling at a loss. That's a far fetched conclusion given the data we have. In any case, like Arun said don't take it personally - any criticism is aimed at your claims, not you specifically

 

[tannat]

I'm not picking on you but I thought you were pretty explicit.



That's a pretty clear statement that you believe some products aren't profitable. Though you contradicted that a few posts later with:



I agree that there isn't enough information to make a definitive claim either way. Though I'm baffled why you think the evidence (45% margins) is in any way supportive of your notion that some products may be selling at a loss. That's a far fetched conclusion given the data we have. In any case, like Arun said don't take it personally - any criticism is aimed at your claims, not you specifically

These quotes are fragments taken out of context and disregarding the whole discussion inbetween, the one Arun tried to point you to. I was thinking from a overall profit perspective from the beginning, while Arun and Silent Guy pointed out that this statement looked like I claimed that products were surely sold below fabrication cost.

I'm quite stunned that you pull this kind of stunt without any visible spirit of discussion, you may feel that you caught me with inconsistency, while I wonder whether you are to busy with positioning.

Anyway, done deal, I don't take it personally, I appreciate discussions. However, you made my possibility (again) become a conclusion in your paragraph above. Or maybe I misunderstood what you tried to say?

That some of the products may be sold as loss is a fact and a definitive possibility. Not a far fetched conclusion. Personally I don't know whether it's likely or not though, but I'm not going to pretend that it's not an open question.

The fact is that we don't know and have no evidence of either.

The notion of several possible outcomes is not a conclusion. Conclusion is left for one of the outcomes.

 

[-The_Mask-]

 

[LordEC911]

Seems to be a similar size to GTX460, similar PCB. Heatsinks on the phases are a nice bonus.

So if this is fully enabled where are we expecting it to fall? Between 6870 and 6950 for maybe $249-$279?

 

[trinibwoy]

Didn't the guys at XS show those pics to be fake?

@tannat, advocating something is possible simply because there isn't conclusive evidence to the contrary isnt a very useful exercise. Especially when the evidence we do have points in the other direction. If your point is simply that "it's possible" then sure but I'm not sure how that helps the conversation.

 

[AlphaWolf]

@tannat, advocating something is possible simply because there isn't conclusive evidence to the contrary isnt a very useful exercise. Especially when the evidence we do have points in the other direction. If your point is simply that "it's possible" then sure but I'm not sure how that helps the conversation.

What evidence to the contrary? There's a whole lot of supposition, no evidence.

 

[tannat]

Didn't the guys at XS show those pics to be fake?

@tannat, advocating something is possible simply because there isn't conclusive evidence to the contrary isnt a very useful exercise. Especially when the evidence we do have points in the other direction. If your point is simply that "it's possible" then sure but I'm not sure how that helps the conversation.

What evidence do point that much in the other direction? I see people claim this but don't see any evidence shown. We have no idea. I've seen some well thought arguments but none that we surely can say may impact - too many ifs.
This conversation started with people using numbers from nvidia as "proof" for one of the outcomes, when truth is that these number did not point appreciably in either direction. This kind of failure of deduction help no conversation. To restate that it is an open question is the rational thing to do IMHO when people jumps to conclusions. - And constructive in my mind.

 

[pcchen]

What evidence do point that much in the other direction? I see people claim this but don't see any evidence shown. We have no idea. I've seen some well thought arguments but none that we surely can say may impact - too many ifs.
This conversation started with people using numbers from nvidia as "proof" for one of the outcomes, when truth is that these number did not point appreciably in either direction. This kind of failure of deduction help no conversation. To restate that it is an open question is the rational thing to do IMHO when people jumps to conclusions. - And constructive in my mind.

Actually, selling a product, especially a mainstream product, at loss, is basically a rare event. So if someone claims that a company is selling a mainstream product at loss (which is rare), he should provide a convincing evidence for that. For example, if NVIDIA is having an operating loss after the said product was release, and having an operating income before, then it can be seen as an evidence.

However, the situation is that NVIDIA is actually having an operating income, with a gross margin higher than its main competition. Therefore, the evidence is basically "against" this rare event. Who claimed that NVIDIA is selling at loss should provide a more substaintial evidence than that. Otherwise, it's safer to assume that the claimed "selling at loss" doesn't actually exist (simply because it's generally rare).

This is basically the same as claiming some guy is doing something rare (for example, claiming some guy is a cross dresser at a bar). You can't just say "you can't never proof that he is not a cross dresser at that bar!" Because if this holds you can say anything because many things can't be disproved.

 

[tannat]

Actually, selling a product, especially a mainstream product, at loss, is basically a rare event. So if someone claims that a company is selling a mainstream product at loss (which is rare), he should provide a convincing evidence for that. For example, if NVIDIA is having an operating loss after the said product was release, and having an operating income before, then it can be seen as an evidence.

However, the situation is that NVIDIA is actually having an operating income, with a gross margin higher than its main competition. Therefore, the evidence is basically "against" this rare event. Who claimed that NVIDIA is selling at loss should provide a more substaintial evidence than that. Otherwise, it's safer to assume that the claimed "selling at loss" doesn't actually exist (simply because it's generally rare).

This is basically the same as claiming some guy is doing something rare (for example, claiming some guy is a cross dresser at a bar). You can't just say "you can't never proof that he is not a cross dresser at that bar!" Because if this holds you can say anything because many things can't be disproved.

You and trinityboy makes it sound as selling products at a loss for limited amount of time to gain/protect marketshare is that unheard of that the possibility should be ignored, or at least not spoken of.

This scenario is not that unlikely, of course only short term. And under the circumstances it may have been a good strategy anyway, if so.

The operating income does include everything and is no evidence for either or neither. The only thing we know that the total operating of graphic cards and chipset together is around 8%.

 

[entity279]

You and trinityboy makes it sound as selling products at a loss for limited amount of time to gain/protect marketshare is that unheard of that the possibility should be ignored, or at least not spoken of.

Well, it's not like aquiring marketshare is a purpose of its own. Why would it be? So I would say that the possibility should at least be ignored.

 

[AlphaWolf]

Well, it's not like aquiring marketshare is a purpose of its own. Why would it be? So I would say that the possibility should at least be ignored.

Two words.
Brand recognition.

 

[Sxotty]

Two words.
Brand recognition.

One word.
shareholders

 

[AlphaWolf]

One word.
shareholders

So you think shareholders would rather they drop a product line, devalue the whole brand and lose marketshare over losing a few cents a unit? Interesting.

 

[pcchen]

You and trinityboy makes it sound as selling products at a loss for limited amount of time to gain/protect marketshare is that unheard of that the possibility should be ignored, or at least not spoken of.

This scenario is not that unlikely, of course only short term. And under the circumstances it may have been a good strategy anyway, if so.

The operating income does include everything and is no evidence for either or neither. The only thing we know that the total operating of graphic cards and chipset together is around 8%.

Did I say that? I said it's "rare." Rare != Impossible. Since it's rare (i.e. less frequent than not selling at loss), the burden of proof lies on the side of the those who claimed it. It's actually that simple.

No one is saying that it's impossible for NVIDIA to sell GTX 460 at loss. However, to say "you can't rule out it with a positive gross margin" basically says nothing. You can easily say that NVIDIA is selling anything at loss (just not everything), or NVIDIA is having too many employees (or too few), or NVIDIA is actually secretly speculating golds to cover the loss, or anything. These have no meaning at all.

Of course one can argue that doing something can be useful at sometime. However, doing something is probably useful does not mean that NVIDIA is necessarily doing that, or even NVIDIA should be doing that. These are very different things. As for the "cross dresser" example, it's like saying "that guy should look good cross dressing, so he must be a cross dresser!" This is simply not an argument at all.

 

[tannat]

Well, let's take it from the beginning of the conversation.

The original claim was that an 8% profit on the whole Geforce line (Including chipsets) proved that no fermi cards have been sold at loss.

This is not a proof (It's not even an indication). It's still absolutely possible that cards have been sold at loss as well as that they haven't. The 8% total profit does not tell us anything about the individual revenue for the products.

Should we accept such wrong statements to be true - just because it anyway can be considered unlikely that any fermi card has been sold at loss at one time or another?

We have only two possibilities here: A and B

A was claimed to be true with a false proof which means that A and B are still possible alternatives. If B is unlikely or not is irrelevant. The proof is still as flawed.

It's very simple logic, And should be pointed out, although it's not worth the flame-war.


To what degree B is unlikely or not is a different question. I don't think many have enough info to make a good assessment on that. There are strong opinions though, so much have I understood.

 

[pcchen]

Well, let's take it from the beginning of the conversation.

The original claim was that an 8% profit on the whole Geforce line (Including chipsets) proved that no fermi cards have been sold at loss.

No. The original claim is that someone said GTX 460 was/is selling at loss. Then someone came in claiming that NVIDIA is selling "everything" with profit. Both at not necessarily true without more substantial evidences.

However, to say NVIDIA is selling GTX 460, which is a mainstream product, at loss, you need to provide a better evidence considering NVIDIA is still having a positive gross margin, which certainly make "selling at loss" a smaller possibility.

Also, there is no "proof" here, unless NVIDIA publicly announce that they are actually doing that. And even with it we can't be sure if what you want is "proof." In real world, many decisions are made with mere possibilities, since many things can never be certain. This type of argument comes surprising frequently though. For example, some people say that you don't know whether gravity works the same on other star systems. Of course we don't know, but based on current evidences, we can say that it's very likely to be the same as on earth. So if you want to say that "gravity doesn't work the same on other star systems!" you need to provide a more substantial evidence instead of just saying "you can't never be certain that a star system in such a distance obey to the same physics rules." That's pretty pointless.