Future of graphics business

[T2k]

Hyperbole.

Excuse me? A whole chain of bad decisions beginning with bad design caused this. It's NV's own fault, they should have accounted for a possible problem with mfring technology - see Anand's long article about how ATI outsmarted NV for years with running new tech on non-important cards to learn the tricks.

Only after the Fermi debacle did NV finally lose majority of DX11 sales for a while. Given the fire sale prices on the GTS 450 and GTX 460, it seems that NV is perfectly fine with fighting for market share by lowering prices.

Hilarious - ATI's next-gen lineup is right around the corner, in 30 days.

Keep in mind NV makes most of its profits not in gaming cards but in things like professional graphics cards where it owns more than 80% of the market and has margins AMD could only dream of.

Well, AMD, you know, this little-known CPU maker, is in black as well but ATI was making money for ~2 years now.

NV will *never* survive exclusively from Quadros, don't be ridiculous.

Does TWIMTBP mean anything to you?

Not much. Only thing comes into my mind is sorry@ss console ports paid by NV and games released literally un-tested on anything else than high- and midrange Nvidia a' la Metro.

NV bankrolls stuff like assisting in creating CUDA programming classes, too.

Yawn... a company makes video tutorials for his products? Whoah, what a surprise. Have you actually ever used CUDA? We use it in-house for about a year now and I can tell you that nobody gives a crap about "CUDA classes", it's not something you will pick up without programming background - our guys learned it here, without going into "bankrolled classes".

Imagine if NV only had 25% of gaming card market share. Would PhysX make any sense? no. If NV had only 25% of the market it would probably give up and join AMD in going OpenCL. Worse, maybe AMD would be the one going proprietary.

Nonsense. AMD or ATI has a history of NOT going proprietary while Nvidia has a long history of ALWAYS ABUSING its market power.
NV is purely a company with bad intentions and corrupt corporate structure, famously overloading the engineers - go and check out glassdoor: http://www.glassdoor.com/Reviews/NVIDIA-Reviews-E7633.htm

Certainly NV took some painful blows like Intel shutting them out of chipsets, the Fermi setback at 40nm, Intel and AMD both going the Fusion route (CPU + GPU, single die) and thus wiping out the low-end discrete card market, doing only so-so in retaining console GPU contracts (the future could be worse; I heard it may be a clean sweep in favor of AMD for the next console generation, ouch), and possibly getting more pricing pressure from Intel should it ever decide to make a sustained push into discrete GPUs again (Larrabee 2).

Man... no offense but you have no idea about this: FYI chipsets were one-third of their revenues. ONE-THIRD.
Mobile market? Apple booted NV for good, Optimus sales are negligible, Tegra is a disaster so far - as you said the only bright spot is their Quadro line but it's obviously not enough to survive when they have to keep pouring billions over billions into R&D, each year requiring more and more.

Ultimately NV may end up a much smaller company, but don't count NV out so long as it owns the professional graphics and can hang even with AMD in discrete gaming graphics. NV also leads in GPGPU (CUDA is easier to program for and already exists). We'll see how NV does in the mobile space, though I agree that so far Tegra-series chips have underwhelmed, financially speaking.

Tesla market is next to nothing, even NV admits it. That's why I'm saying I'm highly skeptical about their long-term viability as a pure graphics HW&SW supplier.

Yes, but NV didn't take kindly to that and is putting pricing pressure on AMD with GTS 450 and GTX 460 price cuts. AMD probably isn't interested in fighting back TOO hard, not when it's still bleeding red ink on the CPU side.

As I said earlier you really need to first learn the facts before posting anything, man...

 

[Martimus]

NV will *never* survive from Quadros, don't be ridiculous.

While I can't say I approve of the tone of your post, I can say that I believe this comment is likely true.

The reason I believe this is that nVidia already owns 80% of the market, so the amount of growth in this market is very limited. In order to increase profit, nVidia would need to make a higher margin on each card sold. This will be very difficult in this market, since they already have very large margins, and increasing the prices would probably reduce the market significantly (since the prices already appear to be near the upper limit that the market will allow).

Since this only accounts for 1/3 of their business, they probably will need other portions of their business to survive to be able to afford development costs. I don't doubt that Tesla will take off somewhat to help this, and I expect nVidia to succeed in the discrete graphics market into the future as well, but I agree that Quadro alone is unlikely to sustain the company (or even pay for its own development).

 

[T2k]

While I can't say I approve of the tone of your post, I can say that I believe this comment is likely true.

The reason I believe this is that nVidia already owns 80% of the market, so the amount of growth in this market is very limited. In order to increase profit, nVidia would need to make a higher margin on each card sold. This will be very difficult in this market, since they already have very large margins, and increasing the prices would probably reduce the market significantly (since the prices already appear to be near the upper limit that the market will allow).

Since this only accounts for 1/3 of their business, they probably will need other portions of their business to survive to be able to afford development costs. I don't doubt that Tesla will take off somewhat to help this, and I expect nVidia to succeed in the discrete graphics market into the future as well, but I agree that Quadro alone is unlikely to sustain the company (or even pay for its own development).

Agreed except I cannot see how would Tesla take off. It's the catch of 22: CUDA to become a success needed to run on *any* NV card and that's why very few companies will buy Teslas for 10x the price.
I simply cannot see Tesla succeeding as a new track along x86/64, not with this pricing structure, period.

 

[blastingcap]

I'll be quick because I don't think I like your tone of voice and will likely not be responding to you after this, in this thread.

1. You said NV's market share disappeared in less than 2 years. Disappeared means zero market share. I called what you said hyperbole. I stand by that.

2. NV is sacrificing profits to maintain market share. It's desperation but it's working for now. If AMD takes its foot off NV's throat by pricing Northern Islands badly, that would be a big mistake.

3. I don't see how this contradicts what I said; in fact I think we agree. AMD as a company barely made a profit despite INTC's big settlement. Judging by DX11, I am guessing that ATI is the profit maker right now, which means without ATI, AMD as a company would likely be in the red. I never said NV can survive with just Quadros; read what I wrote again, the part about hanging even with AMD in gaming cards, etc.

5. You're missing the point. NV's fat profits in the years leading up to Fermi allowed NV to continue bankrolling stuff like TWIMTBP.

6. Again you are missing the point re: CUDA instruction. Note that DirectX beat out OpenGL. Without enough money behind a closed standard... then again I agree with Hurley that CUDA's lead may not be enough.

7. Again you're missing the main point. NV must have a reasonably large marketshare to make CUDA/PhysX worthwhile. If it had 25% marketshare, do you think F@H would be optimized for NV GPUs, for instance? I've actually read NV Glassdoor reviews. Have you read AMD ones? Neither company has fully "gruntled" employees, so to speak.

8. Chipsets are not where the profits were made. If NV really had to, it could lay off everybody displaced by the decline of the chipset unit, but it would be very bad for morale. I think JHH said something like 1000 engineers used to work there. NV saw the writing on the wall though, so I suspect as many of the good engineers were re-assigned to other units. Also many of the engineers themselves probably saw the writing on the wall and left for other units or companies long ago. In business, nothing is ever booted "for good." People tend to not burn bridges unless they must, so if AMD stumbles for a couple of years in a row, NV will become an option again.

9. NV has first mover advantage in GPGPU and although the revenue is small, the possibilities are good to use more and more NV gear in rendering movies like Avatar, or on wall street, supercomputing, or in seismic analysis. It won't be enough but it's a start, and NV MUST continue to push the envelope here because of its business failures elsewhere.

10. I would tone down your condescending and generally antagonistic tone of voice lest Idontcare comes over and tones it down for you. I don't mind civil discourse, though.

Excuse me? A whole chain of bad decisions beginning with bad design caused this. It's NV's own fault, they should have accounted for a possible problem with mfring technology - see Anand's long article about how ATI outsmarted NV for years with running new tech on non-important cards to learn the tricks.

Hilarious - ATI's next-gen lineup is right around the corner, in 30 days.

Well, AMD, you know, this little-known CPU maker, is in black as well but ATI was making money for ~2 years now.

NV will *never* survive exclusively from Quadros, don't be ridiculous.

Not much. Only thing comes into my mind is sorry@ss console ports paid by NV and games released literally un-tested on anything else than high- and midrange Nvidia a' la Metro.

Yawn... a company makes video tutorials for his products? Whoah, what a surprise. Have you actually ever used CUDA? We use it in-house for about a year now and I can tell you that nobody gives a crap about "CUDA classes", it's not something you will pick up without programming background - our guys learned it here, without going into "bankrolled classes".

Nonsense. AMD or ATI has a history of NOT going proprietary while Nvidia has a long history of ALWAYS ABUSING its market power.
NV is purely a company with bad intentions and corrupt corporate structure, famously overloading the engineers - go and check out glassdoor: http://www.glassdoor.com/Reviews/NVIDIA-Reviews-E7633.htm

Man... no offense but you have no idea about this: FYI chipsets were one-third of their revenues. ONE-THIRD.
Mobile market? Apple booted NV for good, Optimus sales are negligible, Tegra is a disaster so far - as you said the only bright spot is their Quadro line but it's obviously not enough to survive when they have to keep pouring billions over billions into R&D, each year requiring more and more.

Tesla market is next to nothing, even NV admits it. That's why I'm saying I'm highly skeptical about their long-term viability as a pure graphics HW&SW supplier.


As I said earlier you really need to first learn the facts before posting anything, man...

 

[sandorski]

NV will *never* survive exclusively from Quadros, don't be ridiculous.

I disagree. They could, but if they did they'd end up like Matrox. A fraction of what they used to be and too far behind to ever get back into the Mainstream.

 

[extra]

I agree with blastingcap that nVidia is, I beleive, the stronger GPU company compared to AMD. But I also beleive AMD knows this and that's why they have their foot on the gas pedal, executing and pushing out cards like no tomorrow.

I would also be scared of Fermi II, nVidia will want to come out fighting with this one.

Ehh, I don't think so as much anymore. Maybe you meant stronger gpu-compute company? I would strongly agree with that statement (that they are stronger gpu-compute wise).

I actually think AMD is the stronger GPU company at the moment if we are talking about gaming.

The truth is despite all the doom and gloom talk, neither one is going anywhere. They've survived this long for a good reason. One bad or mediocre launch isn't going to bury either one of them.

Think about this:
1. For Nvidia, they made the crappy FX series. But then they went on to make the 8800, one of the best cards of all time.
2. For ATI, they made the mediocre 2900, but then went on to make the 4850, one of the best cards of all time.

Fermi was a very mediocre launch, but it wasn't an FX series--it was more like a 2900 series--hot, takes a lot of power, but performance is good.

And I think the parallels continue in that I think Fermi II will be very strong, similar to how the 4xxx series was. But AMD's VLIW architecture is also very strong. They will both be around for a long time--and we need that as consumers.

 

[badb0y]

Ehh, I don't think so as much anymore. Maybe you meant stronger gpu-compute company? I would strongly agree with that statement (that they are stronger gpu-compute wise).

I actually think AMD is the stronger GPU company at the moment if we are talking about gaming.

The truth is despite all the doom and gloom talk, neither one is going anywhere. They've survived this long for a good reason. One bad or mediocre launch isn't going to bury either one of them.

Think about this:
1. For Nvidia, they made the crappy FX series. But then they went on to make the 8800, one of the best cards of all time.
2. For ATI, they made the mediocre 2900, but then went on to make the 4850, one of the best cards of all time.

Fermi was a very mediocre launch, but it wasn't an FX series--it was more like a 2900 series--hot, takes a lot of power, but performance is good.

And I think the parallels continue in that I think Fermi II will be very strong, similar to how the 4xxx series was. But AMD's VLIW architecture is also very strong. They will both be around for a long time--and we need that as consumers.

Actually, I didn't revise what I wrote, what I meant was marketing-wise and brand recognition-wise, nVidia is ahead of ATi. A lot of the average people I have met go so far as to equate nVidia with a graphics card.

Me and BenSkywalker had a debate about this a while back and we both agreed that nVidia has the stronger name in Graphics market and for ATi to really shake nVidia = GPU mentality of the average user they would have to not have 1 generation of successful cards but multiple generations and as far as we can see right now, it seems AMD is doing exactly that.

 

[Dribble]

Why is it that every video thread these days ultimately devolves into NVIDIA's market share or business strategy?

Because a sizeable subset of this forum seems to have very little interest in video cards beyond what it means for company A to be pwning company B (it has to be said mostly ati pwning nvidia).

Some hard moderation would be nice - I can't see why this thread has to mention nvidia the company at all? Sure mentioning I'll replace my current nvidia card X with this one when it comes out is ok, but any mention of nvidia the company going bust, or nvidia the company being nasty, or nvidia's corporate decisions on batman AA/physx/opencl, or whatever can get deleted.

 

[HurleyBird]

If he's confused you are even worse becuase it has nothing to do with "bad drivers" - it's simply no or little tools.
AMD simply does not assign legions of engineers to professional card support, they don't push OpenCL like NV does CUDA etc.
It's economics, simple as 1-2-3 - they first needed to reach profitability and the easiest way to do it on the commodity graphics card market, not on the professional ones.
OTOH if this next 68xx-series will be successful I'm sure as hell they start ramping up their professional efforts especially OpenCL support.

You misunderstood my post. I'm not talking about GPGPU, CUDA, OpenCL, etc. I'm talking about pro rendering, which is the vast majority of the pro market. Again, Quadro, *not* Tesla. In this area AMD's drivers do very much need to improve.

 

[T2k]

You misunderstood my post. I'm not talking about GPGPU, CUDA, OpenCL, etc. I'm talking about pro rendering, which is the vast majority of the pro market. Again, Quadro, *not* Tesla. In this area AMD's drivers do very much need to improve.

But that's what I disagree with...

 

[RussianSensation]

Please, RS - either pont out anything or just remain silent, your response is way too childish, even for the VGA forum.

1. NV's CEO has already acknowledged that the firm does not concentrate on the chipset business at all. Pointing out that many decades ago the chipset business comprised 1/3 of revenue is irrelevant. Even now, the profitability related to this business is barely 10% of profits. Everyone knows that NV's future growth opportunities are:

a) Notebook graphics (such as expanding into adding standalone discrete graphics for notebooks - similar to AMD's Lasso)

b) Professional graphics (which comprise 30% of profits and where NV commands 88%+ market share) such as Quadro and CUDA specific applications for businesses

c) Mobile CPUs such as Tegra2, etc.

d) Discrete Graphics (obviously).

Not to mention that the chipset market stopped being a growth market for NV for years, ever since Intel didn't grant them a license to participate with Core i infrastructure systems.

 

[Viditor]

1. NV's CEO has already acknowledged that the firm does not concentrate on the chipset business at all. Pointing out that many decades ago the chipset business comprised 1/3 of revenue is irrelevant. Even now, the profitability related to this business is barely 10% of profits. Everyone knows that NV's future growth opportunities are:

a) Notebook graphics (such as expanding into adding standalone discrete graphics for notebooks - similar to AMD's Lasso)

b) Professional graphics (which comprise 30% of profits and where NV commands 88%+ market share) such as Quadro and CUDA specific applications for businesses

c) Mobile CPUs such as Tegra2, etc.

d) Discrete Graphics (obviously).

Not to mention that the chipset market stopped being a growth market for NV for years, ever since Intel didn't grant them a license to participate with Core i infrastructure systems.

some issues...

1. "many decades ago" is obviously silly, since the company is less than 2 decades old
2. Where did you get the "profitability related to this business is barely 10% of profits" from?
3. There's a reason that Lasso never developed...

I agree that Nvidia's chipset days are dropping fast, but that's because Fusion and Sandy Bridge will be making the chipsets much more irrelevant for IGP solutions (the bulk of the sales).

 

[Keysplayr]

But that's what I disagree with...

Not entirely clear on what you're disagreeing with. Is it the use of the term "bad drivers"? Maybe he should have said anemic professional graphics driver team resources. No, or too little tools. Whatever the reason, in the end the result reflects the resources put into it.

"OTOH if this next 68xx-series will be successful I'm sure as hell they start ramping up their professional efforts especially OpenCL support."

What is special about the 6xxx series that AMD would suddenly crack down on pro graphics division? AMD would need to infuse gobs of resources (money/personnel) for many years to get their pro division (FireGL) up to snuff.
If AMD's CPU division was doing better, I'd say that might be possible, but as far as I can see, AMD's GPU's are what is really floating the company right now.

 

[Skurge]

Not entirely clear on what you're disagreeing with. Is it the use of the term "bad drivers"? Maybe he should have said anemic professional graphics driver team resources. No, or too little tools. Whatever the reason, in the end the result reflects the resources put into it.

"OTOH if this next 68xx-series will be successful I'm sure as hell they start ramping up their professional efforts especially OpenCL support."

What is special about the 6xxx series that AMD would suddenly crack down on pro graphics division? AMD would need to infuse gobs of resources (money/personnel) for many years to get their pro division (FireGL) up to snuff.
If AMD's CPU division was doing better, I'd say that might be possible, but as far as I can see, AMD's GPU's are what is really floating the company right now.

It's all about what they put their resources into I guess and for a while now, they really haven't had that many. Im sure they made a lot of money with the 5xxx series and if they do the same with the 6xxx series, they might be able to put a little more resources into the software side.

 

[T2k]

1. NV's CEO has already acknowledged that the firm does not concentrate on the chipset business at all. Pointing out that many decades ago the chipset business comprised 1/3 of revenue is irrelevant. Even now, the profitability related to this business is barely 10% of profits. Everyone knows that NV's future growth opportunities are:

The fact remains a fact: when NV lost its chipset business it was 1/3rd of its entire revenue. Period.
No more chipset business, lost revenue, period.
FYI quoting JHH is rather silly here when his #1 priority was to calm investors (let's forget his usual dishonest comments); can you imagine him saying "yeah, it sucks, we lost it despite all our efforts to get a license and save 1/3rd of our revenue"? Stock would tank immediately.

a) Notebook graphics (such as expanding into adding standalone discrete graphics for notebooks - similar to AMD's Lasso)

Totally going down. It's not just their lower-than-average production quality - see Apple's troubles - but to Intel's integrated cores especially in iX-series chips: they are actually pretty decent for everyday computing including video & Flash except 3D gaming.
At the same time AMD was ablo to ship truckloads of mobile DX11 chips so their high-end mobile graphics market share lead also disappeared and now AMD is coming with their intergated CPU+GPU too... future looks pretty dark here, NV has zero chance to beat AMD in this game on the long run, forget Intel's market share...

b) Professional graphics (which comprise 30% of profits and where NV commands 88%+ market share) such as Quadro and CUDA specific applications for businesses

Link to back up that 88% claim? I highly doubt it's true when ATI has truckloads of low- and mid-range WS card, usually for less than NV.

May I ask have you ever seen a CUDA app in your life? Because we use CUDA here and I can tell you CUDA is NOT a real market, stop bringing it up like if people pick CUDA to do whatever they want to do - it's the other way around, when you have something specific task that would be more effective to run on the GPU then you probably went with CUDA, thanks to the lack of tools for OpenCL.

FYI Quadros are all based on the same R&D budget discrete graphics sales finance - if discrete graphics profits disappear so will the R&D budget, along with Quadros.

c) Mobile CPUs such as Tegra2, etc.

Yeah, complete disaster. Billions poured into Tegra, Tegra 2, the smartphone claims - then, as always, it turned out NV is unable to back up the claims, they eat a lot more power than promised etc etc.

d) Discrete Graphics (obviously).

This is where we started: they lost their market leader position to AMD in very short time and does not look like they will make it back anytime soon.
And the future does not look rosy here either: as soon as AMD starts rolling out entry-level and mid-range performance graphics cores integrated into the CPU the bread & butter market start disappearing for NV because they do not have x86 license.
This is why they are desperately throwing money at CUDA & Tesla, hoping to build up their niche market for the rainy days, when desktop graphics will be mostly ruled by integrated products from AMD and Intel.
High-end gaming card market will stay but without making truckload of money on the mid-range and entry-level NV has no chance to keep up with AMD's increasing R&D budget which, in turn, will effect their Quadro line as well.

This is speculation but well-founded, I believe. Time will tell but NV has to either come back very strong in both mobile and discrete segment to pull the rug under AMD or suddenly convince lots of companies to give up x86

Not to mention that the chipset market stopped being a growth market for NV for years, ever since Intel didn't grant them a license to participate with Core i infrastructure systems.

You are confused, IMO: that was the sole reason why NV lost it completely. Otherwise they would be all over it already.

 

[Dark Shroud]

Not entirely clear on what you're disagreeing with. Is it the use of the term "bad drivers"? Maybe he should have said anemic professional graphics driver team resources. No, or too little tools. Whatever the reason, in the end the result reflects the resources put into it.

"OTOH if this next 68xx-series will be successful I'm sure as hell they start ramping up their professional efforts especially OpenCL support."

What is special about the 6xxx series that AMD would suddenly crack down on pro graphics division? AMD would need to infuse gobs of resources (money/personnel) for many years to get their pro division (FireGL) up to snuff.
If AMD's CPU division was doing better, I'd say that might be possible, but as far as I can see, AMD's GPU's are what is really floating the company right now.

It's all about what they put their resources into I guess and for a while now, they really haven't had that many. Im sure they made a lot of money with the 5xxx series and if they do the same with the 6xxx series, they might be able to put a little more resources into the software side.

It's only been in the last year that AMD in general has been in the black for a while. Yes the ATI division has been in the black for some time but that was to the point of propping up the whole company in a way.

What many of us are hoping is that AMD is now able to put some money & manpower behind their software, particularly their SDKs. Up until recently AMD has had to rely on the community in developing software/APIs like OpenCL and such. I personally am hoping they really push & support Bullet.

And maybe just maybe getting a dialogue with MS on adding hardware acceleration possibly Physics in general in a future version of DirectX to make official specifications. Since that seems to be the only way to get things done.

 

[T2k]

<placeholder>

 

[Barfo]

<placeholder>

wat?

edit: Sorry, carry one with the discussion.

 

[zagood]

<placeholder>

 

[Idontcare]

<placeholder>

wat?

<placeholder>

This is a bit of an experiment folks, please be kind as I'm only trying to do my best here.

T2k started this thread after discussion with me in pm, and he has requested I move his relevant posts from the "AMD confirms new architecture" thread over to this thread.

Per my request here, if anyone else would like to have their posts moved from the AMD thread over to this "graphics business" thread then please pm me with the specific URL of your post(s) and I will move it/them here.

The chronology of your posts will be retained...meaning the timestamp you made your post in the AMD will be used for determining where your moved post will go in this thread.

Moderator Idontcare

edit: OK, I think I got most of them. If anyone still wants more of their posts moved between the threads just pm me and I'll get to it eventually. Good luck, and keep it on topic folks

 

[gorobei]

<lawnchair seatwarmer>

 

[BD231]

edit: Methinks a thread should not be started based on a flat out argument :thumbsdown:. A thread stemming from an Anandtech article is inevitable correct? Why not let it start itself so as not to detract people who don't like arguing?

Just my .02

 

[Martimus]

Everyone knows that NV's future growth opportunities are:

b) Professional graphics (which comprise 30% of profits and where NV commands 88%+ market share) such as Quadro and CUDA specific applications for businesses

This statement confuses me. If they already command an 88% market share, where is the growth going to come from? They either need to figure out a way to make the market bigger, get more profit from each part sold, or fight for the final 12% of the market. None of these options are likely to get much in the way of growth however.

 

[Flipped Gazelle]

http://www.xbitlabs.com/news/other/...tion_Market_Begins_to_Stabilize_Analysts.html

 

[T2k]

This statement confuses me. If they already command an 88% market share, where is the growth going to come from? They either need to figure out a way to make the market bigger, get more profit from each part sold, or fight for the final 12% of the market. None of these options are likely to get much in the way of growth however.

Exactly. It might be a solid revenue source but it won't grow, that's for sure unless he confuses Tesla-sales with Quadro-sales.