NVIDIA Reports Financial Results for Third Quarter Fiscal 2010

[blastingcap]

AMD has been selling millions of 5XXX, shipping up to 90% of DX11 cards according to one article....and all they can muster is $2 million in net revenue. Honestly there is some serious mismanagement over there.

You do realize that NV has a de facto monopoly in pro graphics and with Tesla, right? And can split the R&D costs of Fermi across GeForce, Quadro, and Tesla whereas Radeons have to carry most of the weight since FirePro is uncompetitive with Quadro?

It's not necessarily mismanagement, it's more like AMD is getting owned by Quadro/Tesla.

 

[AnandThenMan]

Kinda ironic that amds slogan is "the future is fusion" or some such, yet Intel and nvidia dont share that same outlook.

Nvidia and Intel DO share the same outlook. They just don't openly admit it. Intel is extremely weak on the graphics side, Nvidia has no general purpose processor of their own. So they both talk about their strengths. AMD currently is the only one that has competent tech in both areas, the only question is if they will be able to execute on their roadmap.

Intel will have a DX11 class GPU in 2012 I think, by then Fusion will be rolling right along (barring major missteps by AMD). IMO Nvidia needs an x86 license, I just don't see them being able to grown as a company by relying on ultra mobile and specialty devices. Time will tell though, JHH is a smart man he's defied the odds before.

 

[RussianSensation]

Unfortuantely that meme will continue, even though there is proof that it is not the case.

Although this will be a good thread to link to every time somebody parrots those remarks.

Maybe just like people started to question the importance of FurMark in regard to maximum power consumption measurements after thinking about it, we might see more hesitance of directly correlating profitability with die sizes between companies without looking at the overall company strategy/management as a whole.

Anyway, you need to repeat after me, "Fermi was not built just for gaming", several times

Agreed.

 

[blastingcap]

Nvidia and Intel DO share the same outlook. They just don't openly admit it. Intel is extremely weak on the graphics side, Nvidia has no general purpose processor of their own. So they both talk about their strengths. AMD currently is the only one that has competent tech in both areas, the only question is if they will be able to execute on their roadmap.

If Intel and NVidia ever merged...

 

[AnandThenMan]

As if that weren't enough reason, the AMD/ATI merger (shutting out NV from AMD's chipset market)

To be clear, AMD did not shut out Nvidia from making chipsets for the AMD platform. Nvidia chose not to develop their chipsets further. I'm paraphrasing, but Nvidia said that AMD processors suck and no one wants them, so why make a chipset for a CPU no one will buy.

Nvidia loves to burn bridges and puff out their chest. Recall the bravado from JHH when talking about Intel.

 

[blastingcap]

URL="http://forums.anandtech.com/showthread.php?t=2102625&highlight=motivation+behind+fermi"]Agreed[/URL].

And as I responded already to Mark, NV's GPGPU push long predates Fermi. Glad we're all in agreement here.

 

[Lonyo]

Forget the profits (because they can be affected by Selling, general, marketing and administrative expenses, R&D, Depreciation, non-recurring items like lawsuits, etc.)

Look at the most important #:

GAAP gross margin increased to 46.5 percent from the second quarter's 16.6 percent

Gross Margin = Revenue - Cost of Goods Sold (Cost of Inventory).

This throws the argument that NV is losing $$ on every GPU sold and that Fermi's large die size is killing NV out the window!!!!!!!!!

Not really. That's company wide gross margin. It has nothing to do with GPUs sold in consumer graphics cards.
They could sell graphics cards/chips to consumers and potentially still make a massive profit and have great margins. It's called the workstation and HPC markets.

http://dl.dropbox.com/u/1061446/psb.png
Quadro and Tesla (PSB) have a massive operative income, over 100x the operating income of GPU, despite having 1/3rd revenue.
(Numbers from end of year report last year).

That's where the money is, and that's what NV is building Fermi for. Money, but not from us.

 

[AnandThenMan]

If Intel and NVidia ever merged...

If, but a)JHH will never surrender control of his company, and b)Intel is already on very thin regulatory ice, the deal would never be approved. Intel already owns the majority of the graphics market (sadly), a merger with Nvidia would put them into monopoly territory.

 

[Genx87]

Nvidia and Intel DO share the same outlook. They just don't openly admit it. Intel is extremely weak on the graphics side, Nvidia has no general purpose processor of their own. So they both talk about their strengths. AMD currently is the only one that has competent tech in both areas, the only question is if they will be able to execute on their roadmap.

Intel will have a DX11 class GPU in 2012 I think, by then Fusion will be rolling right along (barring major missteps by AMD). IMO Nvidia needs an x86 license, I just don't see them being able to grown as a company by relying on ultra mobile and specialty devices. Time will tell though, JHH is a smart man he's defied the odds before.

Mobile phones, Tablet, and netbook will outgrow the x86 market by leaps and bounds over the next decade. Nvidia entering into a stagnated market with two established players is a waste of resources. Better to target new markets that have very good growth potential. I am not in the camp that believes Nvidia needs x86 to survive. They can survive just fine feasting on 100+ gaming GPUs, Quadro, Tesla, and Tegra.

I believe keys has shown the expected unit shipments for things like phones and tablets compared to x86. I believe on a per unit basis x86 is already being outsold 3 or 4 to 1. That is only going to go up over the next decade.

 

[blastingcap]

Mobile phones, Tablet, and netbook will outgrow the x86 market by leaps and bounds over the next decade. Nvidia entering into a stagnated market with two established players is a waste of resources. Better to target new markets that have very good growth potential. I am not in the camp that believes Nvidia needs x86 to survive. They can survive just fine feasting on 100+ gaming GPUs, Quadro, Tesla, and Tegra.

I believe keys has shown the expected unit shipments for things like phones and tablets compared to x86. I believe on a per unit basis x86 is already being outsold 3 or 4 to 1. That is only going to go up over the next decade.

What are the margins like for mobile chips in non-x86 systems vs. x86 notebook and desktop chips? I am talking profitability not revenues, here. I'm curious.

 

[Genx87]

What are the margins like for mobile chips in non-x86 systems vs. x86 notebook and desktop chips? I am talking profitability not revenues, here. I'm curious.

No idea. But I cant imagine margins for a CPU are much better unless you have some really strong market control like Intel. AMD has routinely had low margins selling x86 chips. If they didnt have ATI under their belt, I doubt they would ever sniff a profit. Why would Nvidia want to enter into a market like that where it is nearly guranteed to be a loser from a profitability standpoint?

We focus on PCs on this msgboard. But the market for phones and tablets is simply huge and still relatively untapped. The amount of cash required to keep an x86 CPU division afloat and moving forward just seems to be too much.

btw I think the most fascinating offshoot of this is how this whole explosion of mobile devices affects Microsoft. We are seeing legitimate operating systems that can displace or deny Windows OS marketshare. Android has the potential to be huge!

 

[RussianSensation]

That's where the money is, and that's what NV is building Fermi for. Money, but not from us.

Agreed on this point. Discrete GPU space is still very important for them, but it isn't the most important unit of the firm as Charlie claims. All of the units are important because that's what allows the development costs to be reasonable when considered against 3 distrinct revenue product lines that share the same architecture (Consumer, Quadra and Tesla). I will take a look at the K-2 report in regard to profit margins when I get a chance to take a deeper look into their explanation.

 

[Meghan54]

From Gary Dvorchak at The Street:

Graphics chip vendor Nvidia is up 4% in afterhours trading after reporting in-line earnings per share and revenue and guiding to meaningfully higher revenue for the fourth quarter. Earnings per share of 15 cents beat expectations by a penny, while revenue that fell 6.6% on a year-over-year basis to $844 million exactly matched consensus. Revenue was up 4% sequentially, driven by GPU, which was up nearly 6% quarter over quarter. The professional products segment saw a small decline, while the small consumer products segment surged 13.6% sequentially. Gross margin of 46.5% disappointed slightly, coming in 30 basis points below expectations due to an unexpected shift in product mix to the negative.

 

[blastingcap]

No idea. But I cant imagine margins for a CPU are much better unless you have some really strong market control like Intel. AMD has routinely had low margins selling x86 chips. If they didnt have ATI under their belt, I doubt they would ever sniff a profit. Why would Nvidia want to enter into a market like that where it is nearly guranteed to be a loser from a profitability standpoint?

We focus on PCs on this msgboard. But the market for phones and tablets is simply huge and still relatively untapped. The amount of cash required to keep an x86 CPU division afloat and moving forward just seems to be too much.

btw I think the most fascinating offshoot of this is how this whole explosion of mobile devices affects Microsoft. We are seeing legitimate operating systems that can displace or deny Windows OS marketshare. Android has the potential to be huge!

Hmm okay, I'm not sure of the answer, either.

Btw, INTC doesn't just make CPUs, but other things like flash drives, but their margins are stunning: 66 percent, even in a down economy!

http://newsroom.intel.com/community...ntel-reports-first-11-billion-revenue-quarter

AMD has struggled even pre-merger, let alone post-merger, but their margins (CPU and GPU) are ~45% right now, same as NV.

I can definitely see NV more or less giving up on chipsets per se and doing a mobile/Quadro/Tegra/other SOC/non-low-end GeForce strategy and letting ARM fight INTC for low-power processor supremacy. Meanwhile INTC and AMD split what used to be NV's low-end GPU marketshare by merging it into the CPU.

I can see INTC continuing to diversify (perhaps into discrete GPUs--a company like INTC can shrug off a Larrabee defeat or two, and INTC's resources dwarf those of NV and AMD combined) and keeping its margins awesome by eating most of the x86 profits.

AMD has uphill battles all over the map. They seem mostly focused on consumer CPU/APUs (including notebooks, via APUs), consumer GPUs, and server/enterprise systems. Maybe they will actually enter GPGPU and put real effort into FirePro at some point. Maybe AMD tries to re-enter mobile after it so foolishly divested itself of its mobile division to Qualcomm a few years back. Maybe pigs will fly.

If I had to bet the farm on one of them, it'd be Intel.

 

[cusideabelincoln]

we might see more hesitance of directly correlating profitability with die sizes between companies without looking at the overall company strategy/management as a whole.

Given this statement, the following argument:

4) AMD's small die size strategy allows for their GPUs to be more profitable

can simply amended to:

AMD's small die size strategy allows their GPUs to be more profitable for AMD.

Actually, that wording is vague, so re-worded:

AMD's small die strategy allows AMD to maximize their own profits.

 

[Sickamore]

KitGuru’s portfolio of investments is as broad as it is eclectic. We monitor the markets and look for bargains. That involves analysing patterns. For example, while roller coasting from quarter to quarter over the past 2 years, AMD has managed to create somewhat steady upward pressure on its stock price. So what about Nasdaq darling nVidia ?
In the year or so after AMD bought ATI, chaos reigned. Bringing the two companies together was more work than anyone imagined and almost every part of the newly combined company suffered.
Although AMD bought ATI, on the way in it was ATI that had the stronger of the roadmaps. While AMD was still struggling with the original Phenom and considering selling triple core processors, ATI had the 4000 and 5000 series products well underway.
So many people were sacked from the newly combined organisation that communication levels reached an all time low and everyone was signing up for CV improvement courses.
Against that unhappy backdrop, nVidia launched its strongest line up ever with the world’s first DX10 products, the legendary 8000 series.
Almost overnight, nVidia went from being a company that turned over $1 billion a year to being a company that turned over $1 billion a quarter. By 2008, nVidia was hammering in over $4 billion a year.
Since then, the market has become much more competitive.
On 30th August this year, nVidia posted second quarter revenue of $811 million. Although that was up 4% on the same period in 2009, it was accompanied by a loss of $140 million..

Over the past 2 weeks, nVidia’s nerdiest bean counters have been totalling up the results of the GTX460 launch. Looking at the smiles on the faces of nVidia’s local representatives, it seems that Q3 might generate ‘banging results’.
OK, all well and good, so where is the bar? What level of result will nVidia need to post in order to show that it is well back on track?

• Up to $850m: Got to be considered some kind of failure. With the advent of the awesome GTX460, nVidia must be expecting to boost past this figure
• From $850m to $1Bn: Decent result. Not enough to start ordering fireworks, but good enough
• $1Bn to $1.1Bn: Now we’re talking. nVidia posting over $1Bn would show that the company has really returned to winning ways and that money available to (a) add more titles to TWIMTBP and (b) run more advertising
• Over $1.1Bn: Phenomenal. If nVidia turned a miserly $811 in Q2 to a monstrous +$1.1Bn in Q3, with profitability, then you’d have to say that the boys are well and truly back in town

KitGuru says: While the GTX480 and 470 cards were slow starters and the GTX465 was a complete no-hoper, the GTX460 has changed the game considerably. The advent of the GTX580 and the GTX570 and GTX590 variants that will follow, look very strong. Will the Radeon 6970 be able to take the lead and enhance AMD’s share price at the same time? We’ll know soon enough!




http://www.kitguru.net/components/graphic-cards/faith/nvidia-prepares-to-count-its-beans-for-q3/

 

[notty22]

Name recognition is important.
Its why auto makers pump huge money in to Nascar.
Win on Sunday, the public buys on Monday.
Getting press about having the fastest gpu on the planet, gets in peoples minds. A bigger part of Nvidias profits may come from workstation gpu's, but they would not be there, without their presence in consumer cards.


I was reading various stock web sites for Nvidia bits, of course they list AMD related stories as well, this caught my eye.
Take the grabby title with a grain of salt.
AMD Still Gets Bullied By Intel And Isn’t Worth Your Lunch Money

 

[RussianSensation]

http://www.kitguru.net/components/graphic-cards/faith/nvidia-prepares-to-count-its-beans-for-q3/

"KitGuru’s portfolio of investments is as broad as it is eclectic. We monitor the markets and look for bargains. That involves analysing patterns."

Kit-Guru experts conclude: Up to $850m: Got to be considered some kind of failure.
vs. actual
NVIDIA (NASDAQ: NVDA)
-- Revenue of $843.9 million

-----------------------------------------------------

Well this just about solidifies my view of Kit-Guru as a review website. Now they are not only hardware review experts but are also "experts" in technical analysis!! Who would have imagined that the editors possess such a multitude of talents.

It is often too easy to take a snapshot of performance for any company and make general conclusions by being selective based on a piece of historical information, without considering the fundamentals behind the stock price. Any intelligent investor knows it's not just about 'candles' on a chart graph, especially one that takes a small snapshot and makes a general conclusion about any company's performance into the future. Speaking of "track records" of good management, performance, shareholder return, why did Kit-Guru mysteriously or conveniently ignore this elephant in the room:

Based on Kit-Guru's expertise, I am shocked the editors are not working on Wall Street. :\

 

[Sickamore]

I just don't see amd coming at nvidia this time. I believe that 5 series was just a wake up call for nvidia. This time they have fermi right. I will buy two 570 when its out and then go for the new tech.

 

[Aristotelian]

I just don't see amd coming at nvidia this time. I believe that 5 series was just a wake up call for nvidia. This time they have fermi right. I will buy two 570 when its out and then go for the new tech.

What do you base this on? The 6850 and 6870 are selling well, sorry, and another of your smear threads proves this despite the negative spin.

 

[Teizo]

Yeah but did you catch the part where he said Fermi was way above previous-gen geometry and perhaps they overshot? I agree with him (and with AT when it reviewed other cards this year): Fermi's tessellation is overbuilt for current-generation games. Until consoles get in on DX11 tessellation, I doubt that extreme tessellation will matter.

I, for one, appreciate that though because it means that I won't have to buy a new card to replace my 460 for several years which increases it's overall value. As mentioned, considering consoles are holding pc games back graphics wise (though they still look good), this card should be able to deliver adequate performance for some time to come.

 

[blastingcap]

I, for one, appreciate that though because it means that I won't have to buy a new card to replace my 460 for several years which increases it's overall value. As mentioned, considering consoles are holding pc games back graphics wise (though they still look good), this card should be able to deliver adequate performance for some time to come.

Depending on how long "several" years is, you could be right. I mean don't expect miracles here, if you are looking to have it last 7 years. Other things, like shader power, matter as well. 🙂

 

[Meghan54]

Two things bothered me about this current quarterly report.

True, profits went up and that's good. But the GPU segment only saw an increase of 6% while the small consumer segment (Ion, Tegra) rose almost 14%.

And Nvidia's inventories rose by $47M. Rising inventories, which is just product sitting in their warehouses unsold, is not good. I'd have expected inventories to have dropped given the increase in sales per segment.

But honestly, I expected more than just a 6% increase in their GPU segment.

 

[happy medium]

But honestly, I expected more than just a 6% increase in their GPU segment.

You won't see that till next quarter.

 

[Genx87]

I expect their mobile division to continue to see increases. There are a couple handsets that will be out Q1 with Android OS that are being powered by Tegra II. The biggest is probably Motorola's Droid II.