10.16.2007

Intel's Q3 2007 Report

Intel today announced very strong financial results for the 3rd Qtr of 2007:
• Revenue $10.1 Billion, up 15 Percent Year-over-Year
• Operating Income $2.2 Billion, up 64 Percent Year-over-Year
• Record Microprocessor, Chipset and Flash Unit Shipments
• Net Income $1.9 Billion, up 43% from last year, 46% from last Qtr
• EPS 31 Cents (expected: 30 cents)• Gross margin: 52 percent

Even more impressive is the Q4 2007 guidance:
• Revenue: Between $10.5 billion and $11.1 billion. (expected: $10.4B)
• Gross margin: 57 percent plus or minus a couple of points.

The atmosphere at the conference call sounded like it was 1995 once again. Record revenues, healthy margins, incredible growth, laughter and congratulations were in order. Not a word about the competition as if it wasn't even worth mentioning. The closest thing that would remind you of AMD is the comment made by Intel about how they walked away from some of the low-end businesses and having the unique ability to cherry pick markets due to its strong product portfolio.

The global demand for CPUs was healthy and we can expect AMD benefit similarly. To paint a complete picture of the battle for market share between AMD and Intel, we need to watch AMD's margins closely. Market share gains in servers and higher ASP in desktops for Intel means AMD is under pressure from all 3 platforms especially in mobile where Intel's ASPs were also down. We'll just have to wait and see if it's significant enough to cancel the upside in the market.

26 comments:

Unknown said...

Listening to the employee webcast and it is just as upbeat. I wouldn't say 1995, but pre-K8 days for sure.

Anonymous said...

Listening to the conference call:

Cash went up by ~2.3Bil and this is after ~1.3Bil was spent on stock buyback and dividends during Q3... so if Intel was like AMD (no dividend&no buyback) their cash would have gone up by ~3.6Bil in just one calendar quarter.

Desktop ASP's INCREASED. Notebooks were slightly down (attributed in part to overall volume growth in mobile)

"Inventory is lower than what we would like" (I think they said it went down 0.6Bil)

"We are being more selective on business we take...And on the other end we walked on a lot of the low end desktop and notebook" (Otellini)

"people tend to focus on US retail market, but worldwide economy and sales are far more important" (paraphrasing a bit - but keep this in mind next time folks roll out those AMD US retail #'s!)

Might as well leave a few crumbs for AMD... And here was Scientia claiming Intel was soaking up Via's market share a while ago - looks like that may no longer be the case (if it was even the case when Scientia postulated it)...

The other curious note is that it looks like Intel may be capacity constrained - Intel indicated that they would not likely get inventory up to the preferred level in Q4 AND hit the revenue projection (meaning if demand is as projected, inventory will still be leaner than desired). They said something to the effect that they felt they could meet current Q4 customer demand, but if it picks up even more they would be pressed to meet it.

And to me the most comical portion was a question on quadcore mix in desktop space (Intel indicated that this was strong in the >$150 desktop proce segment). The funny thing - they indicated it will be difficult to predict demand and mix in 2008, but for Intel it's just a matter of making more quads instead of duals - I think a veiled reference that because of the MCM approach, Intel has an ~13week advantage on reading the market over AMD as they don't have to decide between dual core vs quad MCM until AFTER the wafers are out of the fab (AMD has to predict this prior to the wafer starting production!)

Yet another example of the business advantages of such a poor design, "glued", inelegant product! (as of course opposed to tripple core!). I'm still waiting for Intel to publish the "quad core for profits" book.

Anonymous said...

Can't wait for Scientia's summary of AMD's plan of victory with an arsenal of underperforming R600 and K10 against the power of CORE and 45nm.

Vrooooooom

JumpingJack said...

I am anxious to see how Intel will lose enough money between now and Q2 08 to go BK after getting fragged :) :) :)

Anonymous said...

"I am anxious to see how Intel will lose enough money"

Come on:

~12Bil in cash on hand/3 quarters = 4Bil loss per quarter...

As Intel estimated 10Bil in revenue next quarter, that means 14Bil in expenses/quarter...

So basically Intel just has to burn a little more than 1Bil every week... or build a new factory ~EVERY MONTH!

They just need to read "how to burn through an obscene amount of cash in a short period of time" for dummies (authored by Hector Ruiz) and no problemo....

Other ideas:
1) They can move to "asset lite" - it's a bit complicated and I don't want to go into the details but it is a very good way of losing money yet keeping the press thinking that you are doing something about it.

2) They can try to become the first company with a native octo-core (excluding cell); while MCM maybe similar and cheaper, "native" octo is better! And while they're at it they could be the first with Penta, Sexta, and Septa cores. We'll be vague on whether this is becuase some of the cores on the outstanding octa design are non-funtional - customer are DEMANDING native Septa-core chips and we will meet that demand! (obviously not all all SW can use 8 cores, so obviously there is a market for SW that can only use 7!)

3) They could buy AMD and payoff al lof their debt! Oh, actually Intel would still have money likely leftover after that - never mind. When you actually think about it - Intel has enough cash to buy AMD outright and payoff all of their debt! (puts things in perspective, no?)

4) They could buy every single chip AMD can produce, throw them in the trash and become a complete monopoly. Again I think they would still have significant cash left over given AMD's quarterly revenues, so never mind on that one too...


You know what the more I think about it, the more I think it is probably not likely Intel will go bankrupt in Q2'08...huh! Who would've thought that that Shari-kook guy would have gotten this one wrong, he's been so good with all of his other predictions.

JumpingJack said...

"Come on:

~12Bil in cash on hand/3 quarters = 4Bil loss per quarter...

As Intel estimated 10Bil in revenue next quarter, that means 14Bil in expenses/quarter..."

I was joking.... the dear Dr. has actually changed his prediction for Intel BK to 2009... He keeps pushing it out actually.

Anonymous said...

5) Put Hector in charge. No amount of money is too big for him to lose, no matter how small amount of time to lose it is.

Axel said...

Copied for posterity from Scientia's blog:

lex

Q3 results for INTEL are in... huge pop in units, stability of price, huge increase quarter to quarter and YoY in profts.

The astounding thing is that Intel claim to have sold over 2 million quadcores in Q3 alone. I don't know what the average ASP of an Intel quadcore is but I imagine somewhere around $400. That means that Intel realized nearly $1 billion in revenue from quadcore alone for Q3.

And there were people claiming that quadcore wouldn't make up much of Intel's revenue base this year? I hate to say I told you so, but in the AMD: Limited Options entry on August 17, Abinstein claimed:

"Quad-core is very small percentage in 2007 desktop, anyway. With the introduction of Yorkfield this is expected to change."

And I responded with:

"No, Intel's guidance on quad-core adoption in 2007 is sure to change in their next earnings call. Already 204 reviews of Q6600 on Newegg, these are selling like mad."

Now the majority of this 2 million was probably server, but the sheer size of the number clearly indicates that Intel's strategy is to push quadcore into the mainstream in order to strangle AMD on capacity. Intel evidently have the capacity to do it and are succeeding quite well at this strategy.

anonymous said...

Just in case anybody couldn't find their calculator:

Gross profit this quarter was a bit over 5.2B. With the projected increase in margin this quarter, the low end of range on revenue and GM yields almost 5.8B in gross profit. High end of both, 6.55B.

In other words, the kooks of the world need to factor in to their bankruptcy calculations that Intel is looking at another 600M to 1.25B in gross profit next quarter. Wait- I know! Intel will go broke paying for fuel trucking all that money back to Santa Clara :9

Anonymous said...

It must be monopoly profits.

Tonus said...

Intel may drive quad-cores into the sub-$200 space by mid-2008?

http://www.overclockers.com/tips01231/

"Q. At this pace, quad-core is in a mix of desktops. What do you expect the mix to be a year from now? Right now it's probably in the low single-digits, low double-digits, as a percent of your desktop shipments?

A. Well, that's a demand question. The supply can be almost anything we want it to be. So it really is a function of how the demand grows at certain price points and what the competitive environment for quad or for things greater than 2 core are. Right now we see quad tending to populate the price points for microprocessors that are sort of 150 and above, and that tends to be a double-digit percent of the market but not 50% of the market."

Anonymous said...

Axel - in all fairness to Abinstein (who I usually violently disagree with), hs comment states desktop. Intel's claim of 2Mil quadcores includes server and desktop.

There is no breakdown, but I would imagine the vast majority of those 2mil remain server related parts (though desktop is probably starting to grow)

So your 2Bil is off a bit - I still think with the quad core pricing on desktop it will start to impact revenue this year - but keep in mind Intel will take in roughly 40Bil total this year, so even 1 Bil on quadcore desktop is still rather small (~2.5%).

You just need to get used to the big #'s - for AMD 1Bil would be a monster #, for Intel it is change they found under the sofa cushions in the lobby at HQ in Santa Clara.

Anonymous said...

Ok, enough rant for me, just for moment. INTC is doing very well. AMD has effectively lost the war and 45nM and 32nM is a lead pipe cinch. The stock is postured to go to 34. My computer, well I’m never satisfied, but I will say I’m comfortable with the Q6600, before Yorkie, of course. The X1900XTX and CrossFire setup is turning 11350 in 3DMARK’06, not bad. So, what’s my problem?

Back to the rant, Nvidia, that’s my problem. MY Crossfire setup is getting rather long in the tooth, you see. When ATI was a separate entity, they focused their resources on their products. They sometimes equaled or surpassed Nvidia offerings. Now, ATI’s resources, and personnel, (which is why there’s a mass exodus of key ATI key people) are playing second fiddle to the needs of AMD’s Barcelona flop, as the graphics end of things are generating much needed revenue. It will get worse as AMD will, without doubt, throw good money after bad to try to tweak the K10 core for salvation. The graphics end, however, has and will suffer. As it stands, there is really no real competition left for Nvidia. (And I can’t have SLI on my new machine!) If there were ever a new target for INTC to set its sights on, I’ve got a big fat juicy ONE, NVDA!

(For the record, NVDA has reached record highs, this date, as they merrily fleece the enthusiast market unchecked.)

Let me say that again, I can’t have the fastest 2 card setup because Nvidia wants to sell chipsets. But, I like Intel chipsets. NForce chipsets have had their issues, not to mention motherboard driver problems. Also, I find INTC chipsets to be more stable. What’s ironic is how many more graphic cards Nvidia would sell IF they opened up SLI to INTC.

AMD has effectively shot itself in both feet. They dropped the ball with Barcelona and ATI is dieing a slow, low end, death. (God, that ATI buyout was SO STUPID!) ATI’s CEO Horton saw this coming, big time. Nvidia idles back, while having a clear shot on monopolizing the high end graphics market as INTC beat the brains out of AMD. ATI’s position is death by attrition, NVIDIA’s, is success by default.

Hang Sin is a shrewd, opportunistic, son of bitch. He nips at INTC heels and runs away laughing, raking in $900 graphics cards, specifically targeting the enthusiast level. It’s as if he’s single handedly denying Intel performance champion machines. These machines are reserved for Nforce, Nvidia boards with SLI. He is posturing Nvidia main boards as the absolute performance product. This high end, high priced graphic card, price fixed market MUST come to an end and INTC has the chops to do it!

Then I had a dream. Horton and all the refugees from the ATI, plus all the 3D gurus Intel acquired from 3D LABS, quietly go to work for INTC. Oh the joy! This may be far fetched. Maybe I’m in fantasy land, on space station 22, circling Orion’s belt, you know, way, way out there. But, just think of the possibilities!

Horton signs up with Intel to head the graphics division! Wouldn’t it be wonderful to see Nvidia start squirming in its pants, peeing itself if INTC ever really got THAT serious about powerful graphics? Think of it, all these guys who HATE NVIDIA’s guts get unlimited resources to beat the tar out of NVDA! The absolute end to VERY expensive graphics cards would be a reality as INTC could offer VOLUMES of competitive product at 2/3 the price. I /We could then tell Hang Sin to stuff his SLI up his greedy ass.

His chipsets would be rendered superfluous as INTC could fulfill WRECTOR RUINZ grand dream of a truly powerful, fully integrated, unified platform. Leap Ahead.

Oh, God, how I wish. (Sigh)


SPARKS

InTheKnow said...

Oh, Abinstein makes it so hard not to call him on his ignorance. He has got to be the most effective blogger I've ever seen. His ability to generate comments is phenomenal. His latest is classic.

... so in your opinion which company has the better process technology, Intel or AMD? Will the company with better process technology attain better or worse die area efficiency?

His inability to separate process capability from design decisions is mind boggling. The difference between Intel's design philosophy and AMD's was aptly summed up by Guru. Intel designs with the end in mind, to maximize the process window (a robust process if you will). They knowingly leave some efficiency on the table to do this. AMD designs elegant engineering solutions. Their notoriously slow ramps to good yields/high performance show the consequences of their design philosophy.

But to claim that die area efficiency (is that AMD's latest comparison point?) is a consequence of inferior process capability is laughable. There are just too many other possible explanations to be able to say that this supposed yardstick proves AMD has better process technology than Intel.

But as a self-anointed authority on all things semi-conductor related, he can't be educated. What a waste of a mind.

Ho Ho said...

Commenting on the discussion of software level compability and core architectures I wonder when will Abinstein say that VIA CPUs are also very similar to Core/K10 as they also run the same x86 code.

I simply can't believe that he is actually trying to say that when CPUs run the same code they must be having very similar internal architectures. And yet he does say that.

Anonymous said...

Sparks... have you not heard of the Larabeee development (Intel's discrete graphics chip)?

This is much further along then most people think from the HW perspective (I know some folks working on the cooling solutions); not too sure about the SW/driver schedule.

This should hit high end solutions in 2008... not sure when it will hit the desktop market (maybe 2009). From what I understand it is a matter of how well Intel puts together the drivers.

So no competition from Nvidia for next year or so, but they best not pull an AMD K8 - get fat dumb and happy and get steamrolled by a graphics version of Core 2.

Tonus said...

sparks:
What’s ironic is how many more graphic cards Nvidia would sell IF they opened up SLI to INTC.

I'm not sure what you mean by this. I just purchased an Asus board with the Intel P35 chipset and it has two PCIe 16x slots for use in SLI configurations. Is this what you are referring to, or is it something else?

Unknown said...

ho ho
I simply can't believe that he is actually trying to say that when CPUs run the same code they must be having very similar internal architectures. And yet he does say that.

Actually, he said every x86 CPUs share 99.9% of similarities....

I can only surmise that what he meant by "same architecture" is by... they are all called "processor"?

Unknown said...

I'm not sure what you mean by this. I just purchased an Asus board with the Intel P35 chipset and it has two PCIe 16x slots for use in SLI configurations. Is this what you are referring to, or is it something else?

You may have two PCI-E 16X slots, but you can't run SLI on that board. Only boards with nForce chipsets (at this stage) can run SLI.

Tonus said...

giant:
You may have two PCI-E 16X slots, but you can't run SLI on that board. Only boards with nForce chipsets (at this stage) can run SLI.

Ah, I did not know that. Thanks!

Anonymous said...

YOU GUYS MUST WATCH THIS!I was unable to stop laughing for quite some time.

http://www.amd.com/us-en/assets/content_type/DigitalMedia/AMD_Leslie_Triplecore_196K.wmv

The best parts:

"Native, true triple core" (as opposed to 'native, non-true(?) triple core' Is it the most fastest chip too?

"Triple core shows our ability to be able to meet customer innovation through our design excellence and through our manufacturing capability"

I tried listening real close to the last word - for a second I thought it was "incapability", but then I realized I was mistaken!

It actually is AMD's manufacturing capability that enables triple core - had they been able to manufacture quad cores well then there would be no triple core, so in a literal sense it is AMD's manufacturing capability (specifically the lack thereof) that does enable triple core!

Do you think the folks that spout this stuff out actually have to shower after these interviews? On a serious note - I know you have to put out the company line, but is there not a point where PERSONAL INTEGRITY kicks in? I wonder how far over the line before you get to the point of "I can't say that, it is simply not true"? (Do you ever get to that point?)

Anonymous said...

Forgot one other inconsistency in the AMD logic in the video...A lot of applications (in early 2008) will not be able to take advantage of quad core...

Good logic - give people what they need at the best price/performance ratio, right!?!

Well what did AMD say when AMD64 was released?!? What was the marketing spiel then? At the time AMD released desktop AMD64 chips the software that could take advantage of 64 bit was ubiquitous no?

What you mean, it was all about "future proofing"? Well surely they released some 32bit versions which were more price/performance conscious for those who wanted max benefit? They didn't?

So let's review: Today - well some folks should buy triple core because SW (today) doesn't take advantage of quads that much. 4 years ago, buy 64bit SW because SW will EVENTUALLY use it (which by the way is still not even close to
mainstream in desktop space)

Would you like some more losses to go with that cup of hypocrisy?

Anonymous said...

AMD #'s don't look too bad... $400Mil loss (226Mil operating loss).


Margins were up to 41% (a very good 8% gain).

Cash is ~ flat at ~1.5Bil (though I think AMD sold the rest of Spansion off this past quarter to offset the loss and thus keep cash flat?)

Anonymous said...

Tonus, for a second there you had me thinking Alzheimer’s was setting in. Actually, for once, I was hoping I was wrong. Naturally, Giant, the other performance nut on this site filled you in. (I’m willing to bet it bugs him, too. What say you, G.?)

This SLI thing has been an outright slap in Intel's face since its inception. As a mater of fact, there were rumors circulating months ago about SLI running on X38 before its release. NVIDIA vehemently denied it with religious fervor at Computex. In fact, they seemed annoyed that the rumors were taken seriously.

Fortunately, however, the 2900XT’s @ $500 a pop are turning in respectable benchmarks. These are the ones with 1 GB of DDR4 and the GPU is clocking in @ 725 MHz. If AMD can survive INTC present and future onslaught, and if they channel some money into some refinements, a Cross Fire setup might (with a INTC CPU) break into the 18,000’s in 3DMARK ’06. Whereas, a full blown Ultra SLI setup can hit 19500 (O.C.'d Qx6850), albeit at a hefty, cool, TWO GRAND!! Bastards!

There it is, give up 1500 to 2000 points to save a thousand dollars. It looks like that’s the way I’m going to go. But, it still pisses me off. It business, and NVDA knows how to play hardball.

SPARKS

Unknown said...

Tonus, for a second there you had me thinking Alzheimer’s was setting in. Actually, for once, I was hoping I was wrong. Naturally, Giant, the other performance nut on this site filled you in. (I’m willing to bet it bugs him, too. What say you, G.?)

Right on. My P5B Deluxe has two PCI-E slots for graphics cards, but I can't run SLI. There's no technical reason that P965/975X/P35/X38 can't run SLI other than Nvidia locking SLI to the nForce chipset. Right now the only mutli-GPU setup I can use is Crossfire. At least this is one thing that AMD/ATI got right, allowing users to use Crossfire on other chipsets like Intel.


Fortunately, however, the 2900XT’s @ $500 a pop are turning in respectable benchmarks.


I disagree. The 2900XT has awful performance when Anti Aliasing is enabled. With a high end graphics card I'm using AA at all times. I didn't really want to buy Nvidia, but Ati really had nothing that was a decent upgrade from my X1800 XT 512MB card, so I bought the 8800 GTS 640MB. No doubt it's a killer card for the $$$. Nvidia makes great GPUs, I don't doubt that at all.

This SLI thing has been an outright slap in Intel's face since its inception. As a mater of fact, there were rumors circulating months ago about SLI running on X38 before its release. NVIDIA vehemently denied it with religious fervor at Computex. In fact, they seemed annoyed that the rumors were taken seriously.

If you look at an AMD system, nForce has widespread adoption because it's probably the best chipset for an AMD CPU. But on the Intel side we all know that Intel makes excellent chipsets, with stability and performance that are second to none. The only real feature they are missing is SLI support - SLI support is the only reason people buy nforce on Intel. If you could run SLI on Intel chipsets I'd imagine that demand for nforce on Intel would dry up pretty quickly!

There it is, give up 1500 to 2000 points to save a thousand dollars. It looks like that’s the way I’m going to go. But, it still pisses me off. It business, and NVDA knows how to play hardball.

The 8800GT is due out at the end of the month. It's supposed to have most of the performance of the Ultra at a $249 price point. Two of them and a 680i SLI motherboard will be cheaper than the two 2900 1GB cards.

Yes, Nvidia can play hard. The truth is, when dealing with ruthless companies like Intel, they have to.

Ho Ho said...

sparks
"Fortunately, however, the 2900XT’s @ $500 a pop are turning in respectable benchmarks."

You mean in the games where they work and without using AA?


"Cross Fire setup might (with a INTC CPU) break into the 18,000’s in 3DMARK ’06."

Yes, R600 is an excellent GPU if your main game is 3D mark as they are rivalling even GTX there. In real games where you can actually move around they are way behind.


giant
"At least this is one thing that AMD/ATI got right, allowing users to use Crossfire on other chipsets like Intel."

Wasn't it because Intel bought CF license from ATI?


There were modified BIOS'es and video drivers that allowed to use SLI on some Intel chipsets. I remember it was possible on my 975x.

I as an average user who doesn't buy the latest and greatest GPUs find that dual-GPU setups are pointless. They mostly have a point with very high end GPUs. If you buy a slower one in hope to increase performance later by getting another one it is usually better to just sell the old and get a new and faster one.