I want to start by saying this is NOT GOOD FOR AMD.
http://www.digitimes.com/systems/a20061024PR211.html (this link now needs a log in)
http://dailytech.com/article.aspx?newsid=4706
If you look at my posts/comments over the last few days I have been saying Dell is dragging AMD down the price stack and squeezing them on pricing thus hurting margins. I am now convinced this is what is happening. I know I said yesterday that AMD needs to do some discounting on notebooks to gain entry into new customers to break Intel's stronghold in the notebook segment but this sounds like they are going to start hurting themselves.
I repeat, if AMD was not running tight on capacity then it's ok to do these kinds of deals to keep the factories loaded. But considering the reports that they are starving teh channel on desktop parts for the last few months and now the TW ODMs are saying AMD may be short of mobile parts, surely they could have made some of these a higher bin and sold them for better margins at some of the other customers they are gaining entry into. Dell has no strategy and is heading into a death spiral of becoming commoditized but it's tragic they are dragging AMD down. As much as I think Intel is going to gain the momentum for the next 6-9 months, I think AMD is now throwing away the strategic leverage they have built so successfully in the market place with a good range of customers through these deals with Dell.
UPDATE
Dell's new AMD based notebooks are not sub 500$. In fact, a quick spec comparision shows they may actually be more expensive than an equivalent Centrino Duo. We probably need to wait 1-2 quarters to really see how this plays out.
Subscribe to:
Post Comments (Atom)
32 comments:
Dell is a desperate company playing the market share game in the expense of profit margins. AMD cannot play this game with them with their massive debt. With the lack of competetive products from AMD, their Turion is now competing with Intel's mobile Celerons. What a pity.
Of course, DELL/AMD notebook is bad for AMD but DELL/Intel notebook is good for Intel.
So 20million AMD CPUs sold to DELL is a horrible thing!!!
Even 1$ earned on CPU means profit.
Tommorow I'm going to buy Acer notebook with Turion processor. Why? Because it costs only 25$ more than Celeron machine and is 200$ cheaper than Core 2 Duo crap. 64-bit CPU with 3D accelerator.
It is Vista ready, not like Intel crap. Think about it...
Ashenman:
Again, if the problem were this simple, Dell would be fixing it already, and AMD would be completely avoiding it.
Dell does not know how to fix their problem because they are not a technology company. They are a supply chain company. They do not know any other strategy outside of cutting costs and I now think Rollins is just incompetent. Often, people on the inside fail to see the most obvious things. Like Intel's stupid decision to extend the Netburst architecture and skip a generation. For which I think both Otellini and Barrett should be canned.
As for AMD - they needed to win Dell...preiod. Unfortunately, when you work with Dell you don't pick when, where and how. This is the simple truth. Let me contextualise this. Intel has given Dell everything they want and more for 20 years now. Yet, Dell is pissed with Intel because they no longer give them preferential pricing on desktops and are sucking up to Apple. I'm not saying Dell doesn't need AMD in it's line up - they do. But some of this is just plain spitefulness - trying to send Intel a message that if you screw with us, you will regret it. Unfortunately, Dell is screwed themselves...or rather screwing themselves.
As for response to notebooks...we'll have to wait and see. I think selling 500k sub 500$ notebooks will not be a problem. The issue is what it does to AMD and Dell's margins. Dell is single handedly tanking the PC industry. It's not a winning model. There are cases around the world where this has failed. e-Machines in Europe - now acquired. PCL in India - #1 PC company in the 90's...drove themselves out of business. If you read the earnings transcript carefully you notice when Glen Yeung is discussing inventories the issue of the right mix being available to the right customer at the right time creeps in. Now...it just doesn't make sense to me that the channel would be crying for product but instead of supplying them AMD is holding it in inventory. I'm pretty sure that incremental 60 million was not planned for and they will either dump it cheap or write it off because it's not the right product.
Now before you jump down my throat on Intel inventories let me comment. I already said when I was forecasting this qtr that at some point Intel would have to take a write off and they did...100 million. Nothing to be proud of. I suspect they may take another write off in Q4 or Q1 on Netburst parts. Or maybe that's what they're lowering the price on Oct 22. So they're not sitting pretty either. However, AMD's inventory is a surprise while Intel's is not because they clearly disclosed that's what they would be doing. And this is the point. The lower margins, the inventory build up - they are indicators of the problem, not the problem themselves. The reason for this is they are contrary to AMD's strategy to move up the value and pricing food chain. My guess is Hector is in a rush to reach 30% market share because at that point he gets scale - not just manufacturing but sheer size. Which makes his customers dependant on him which means they will not want him to go out of business or lose significant share.
Anonymous said...
Of course, DELL/AMD notebook is bad for AMD but DELL/Intel notebook is good for Intel.
So 20million AMD CPUs sold to DELL is a horrible thing!!!
Even 1$ earned on CPU means profit.
Tommorow I'm going to buy Acer notebook with Turion processor. Why? Because it costs only 25$ more than Celeron machine and is 200$ cheaper than Core 2 Duo crap. 64-bit CPU with 3D accelerator.
It is Vista ready, not like Intel crap. Think about it...
Dude - your body language is not polite. And your argument is pretty senseless. If you have something intelligent to say, please do. If you want to make incoherent statements with bluster and try and force your opinion, there are other places on the Internet you could spend your time better.
Now get your head out of the place where the sun don't shine and explain to me how if AMD sold 30-40% of their capacity for 1$ profit per part on cost of manufacture/sales then how would they make money and what their gross margin would be?
180degree:
One thing you (may) have potentially ignored is that AMD is taking a calculated gamble that their capacity will ramp significantly and while they will take a 1-2 quarter hit on margin and profit, they will be able to make the proift back long term through increased volume (although I suspect margin will be tough to get back unless prices go back up).
The risk here is that they've likely established a bottom dollar price at Dell on both desktop and mobile - if they cannot sell in volume (either through poor execution/yield issue on 65nm or if Intel decides to further the price war) they will be in trouble as they need cash flow for merger costs and for F30 conversion, and NY fab.
I also think they are betting on the fact that when some of their capacity issues ease in Q1-Q2 next year, the channel will come back which should help their margins bounce back a bit... not sure if this is true or not though (depends on how much Intel caters to the channel over the next 2-3 quarters).
Another thought - I put this on your arch-enemies site before the Dell notebook articles came out
I hypthesized that perhaps AMD's mobile unit volume went up so much (50%) yet market share was up only a bit (4%) was because they may have shipped a bunch of parts to a customer (like Dell) which had not made it to market yet - which would therefore not be reflected in market share #'s yet.
Given Dell's propensity to require large volumes - they may have bought a bunch up front (with some sort of volume discount deal).
Any thoughts? If this is true, AMD has already booked most (at least a "good chunk") of the revenue for these upcoming Dell notebooks and the impact of this on AMD's Q4 revenue may be less than people think (and minimal impact to gross margin). Although market share #'s would go up in mobile sector.
Why is such a big deal being made of this? Dell already sells sub $500 notebooks.
Not to put words into anonymous's mouth, but what I think he's trying to say is that of course this is bad news in your eyes.
On the other hand if Dell had intro'd AMD laps and not under cut C2D you would say "BAD NEWS FOR AMD." In fact it's exactly what they've been doing already, Turions have under cut Core offerings by $100-$200 for the past 4-6 months. The only difference was Dell wasn't selling them. So your arguement has to be Turion is just as good of chip as C2D. If not it should be priced down, then your arguement is just semantics. How much should it be priced down? AMD feels comfortable pricing it down this much, you don't. We'll see who's right, but in the end AMD won't be undone because they priced a chip $20-$30 high or low but whether or not they can answer Intel's offerings all while increasing capacity.
Channel are folks who buy through distribution and not direct from AMD or Intel.
They may be re-selling themselves or assembling PCs...or perhaps even both. A significant number of these guys are in emerging markets.
A couple of things:
To Anonymous who posted bout AMD's potential strategy...great post. I think you're right on what they're trying to do.
On the question of whether AMD would have booked the notebook revenue. It depends...on ship dates and cancellation windows. Knowing how Dell manages their inventory, I suspect Dell will click "Buy" as close to build date as possible. Hence, they probably have booked some but not all. In some cases, cancellation windows can be less than a week from ship date.
Ashenman - yes, that's correct. If you want large deals in one shot, you do that through OEMs (Original Equipment Mfrs - i.e. HP, Lenovo, Dell). But if you want to move new technology specially in the consumer/SMB space it's channel that does this. Channel is app 30-40% of the global PC market.
System Integrators are channel who sell more than boxes...they sell solutions like networking, security, maintenance, etc.
Dell has no strategy and is heading into a death spiral
I've tried to substantiate this but I haven't been able to.
Presumably, if Dell were going downhill their margins should be dropping, their revenues down, and they should be running out of cash.
Dell's cash position is up from where it was a year ago. Dell has $6.8 Billion in cash. They spent 1.1% of their cash since last quarter and 3.4% since two quarters ago. However, they are down in cash only 0.6% of where they were three quarters ago. Cash spending does not support your conclusion.
There is no clear trend on Revenue. Dell has been up and down in the past several quarters. If we take a four quarter average then Dell is up 8.8% since last year.
Margins have fluctuated even with the past four quarters. The best we can say is that margins are below where Dell normally is.
For actual margin revenue averaged over four quarters we can say that Dell is down 1.8% from a year ago.
The best that you can say from these figures is that Dell is down slightly from last year. However, the data does not support a "death spiral" theory.
The sub $500 laptp will kill Intel's Core Duo sales. If you look at Intel's 3Q06 results, it was mobile which saved its day. Intel's desktop+server showed almost zero sequential growth (and of course massive y/y fall). So, for AMD to kill off Intel, it must deny Intel's the oppurtunity to get $150 for each mobile CPU. Every $50 AMD sempron sold is $150 less for Intel -- AMD makes $10.
Expect Intel's mobile revenue to drop significantly in 4Q06.
Here we go again. Vista will be shipping in 32 and 64 bit versions. Most CD notebooks can be upgraded to C2D so they can accept either 32 or 64 bit windows. Or you can have a 32 bit Vista install on C2D, your choice.
The chief move to 64 bit is to support applications that need more than 4 gb of memory, per Microsoft. That Microsoft and others are optimizing their code for 64 bit is merely an added bonus.
Those of use that were around for the 8 to 16 bit transition, then the slide from 16 to 32 really ain't all that impressed with 32 to 64. It's just another rung on the technology ladder.
This doesn't belong here, but I'm sure Sharikou wouldn't post it.
In his most recent article he attempts to draw parallels between AMD shorters and Bin Laden. Not that Dr. Sharikou is foreign to shorting stocks, Intel in particular. Oh, the hypocrisy. His entire blog is meant to support his financial bet.
Posted by: Dr. Ge on February 23, 2006 2:13 pm | In Long Ideas, General thoughts, Short Ideas | | E-Mail This Post/Page
My friend Sharikou left some comments on the blog “My Stock Plays Update” concerning my long positions in Intel.
“I have made large gains by taking bearish position on INTC, and am increasing that position. I project that AMD will exit 2006 with 40% market share, and INTEL will suffer substantial losses in 2007.”
Doctor:
Intel grew QoQ in servers and probably lost a little market share in mobile. But as I've had to point out your errors in the past so I'm not surprised you allow a little something like the facts to get in your way.
Remember - Core 2's are cheaper to make than Netbursts. And I have already shown you a couple of times how your calculations are wrong because the data it is based off is wrong.
Ashenman:
1. Core 2's are cheaper - what this means is as Intel moves from 5% mix in Q3 to 25%+ mix in Q4 - their margins improve which means they generate more cash which they can either bank as profit or use to sustain the price war or invest elsewhere. Having a better product which is cheaper to manufacure improves their business financials. Going from almost 100% Netburst mfg to 75% and then to 50% and toward 0 makes Intel's position stronger. At the same time, they continue to increase their capacity on 65nm. So, while AMD starts ramping 65nm to help drive their costs down Intel is doing that but also ramping a mArch that is cheaper to mfr.
2. On the 500$ notebook issue - this will definitely hurt Intel. But I am less concerned about Intel at this time and more on the medium term impact to the tech industry. This will further accelerate the commoditization of the notebook category which means potentially a system level price war among OEMs resulting in lower margins, lower profits and all the bad things that come from being in a commoditized business. I actually hope Dell fails to sell these like they did with the low end Celerons last year because this means a lot of pain for the industry and the people who work in it (i.e. jobs lost) if the other OEMs also decide to focus her to combat Dell.
Ashenman:
1. Commoditization - I'm sorry but I disagree with you. Commoditization by definition means a lack of differentiation between products thereby being differentiated only on the basis of price. Commoditization means no innovation and in the tech business that is not a good thing. It's ok if you're selling potatoes but not here. When margins are wafer thin - the money to invest in R&D for the OEMs disappears. Right now wafer thing margins on desktop are being buffered by notebooks because this segment is growing at around 30% with an average selling price still close to 1000 USD.
2. My mix comment was on desktop btw. Core 2 is the desktop brand and the point I was making was desktop is still the lions share of Intel's volume.
I'll read Scientia's blog in details (gave it a quick glance) but once again I want to reiterate, I am talking about momentum shifting. AMD is not growing at the same pace it has for the last 4 qtrs. Also, Scientia is working with incomplete info. He is assuming that Via's share loss resulted in exactly the same kind of low end business being won by Intel & AMD in proportion of their share gain. Unfortunately, until we see WW market data from Gartner and IDC 3-4 months from now this cannot be verified. Nobody knows for a fact that the 4 points of share Via lost were replaced by low end deals and nobody (including Scientia) knows for a fact that it wasn't either AMD or Intel who picked up 4% of low end while losing higher end SKUs to their competitor. Hence, while Scientia may believe his hypothesis...there is a possible hole in his reasoning. Which is exactly why I am not sticking my neck out to say Intel has won. What I am saying is it is a signal they are re-gaining momentum when they gain share instead of losing it.
Scientia - I guess you're reading this. So let me ask you like Ashenman did to call Intel & AMD's Q4. It's easy to say what is not and defend that position...it's a lot harder to say what it will be. BTW - it would be nice if you didn't take my comments like the one on death spiral out of context. I said Dell is heading into a death spiral of commoditization and you chose to grab only part of my statement.
180 sharikou:
When you say so definetively that C2D is cheaper to make than P4, are you comparing it to 90nm Netburst, 65nm Cedar Mill/Presler when Intel just started 65nm ramp, or their currently produced 65nm Netburst.
I ask simply because there's a heck of a difference. I find it hard to believe that a newly designed chip, albeit derivative, with 4MB of cache can be cheaper than an older chip on a well sorted process. If you were to add the cost of 65nm transition to Netburst, than I could see how what your saying is true. Beyond that I'm pretty skeptical. As Ashenman's arguement goes, why ramp C2D slowly(and it is slow) and continue to produce so many P4's(to go along with the ones sitting in storage) at greater expense when their both on the same process. Any proof otherwise?
Here's some perspective for those obsessing over AMD's margins, without mentioning names of course.
http://www.overclockers.com/tips01055
I can't say it better than Ed.
Randy - die sizes on C2D are significantl smaller. In fact, they're even smaller than Athlon. I dealt with this somewhat here:
http://sharikou180.blogspot.com/2006/09/deja-vu-2002-all-over-again.html
There is a handy Tom's Hardware link in the article.
You're right that there is a cost impact as you ramp a new product into manufacturing but my comment was as C2D (actually Conroe) becomes a larger volume of Intel's mix, their relative costs reduce and hence the money they make per CPU goes up. It's a simple statement...not meant to mean any more than what it says.
Ashenman - two things on Via and Intel competing:
1. Via makes a range of products over and above "Embedded" which is what you're referring to.
http://www.via.com.tw/en/products/processors/
2. Intel also competes extensively in the Embedded business. And that includes derivatives of even Core 2. Check them out here.
http://www.intel.com/design/embedded/
http://www.intel.com/products/embedded/processors.htm
http://www.intel.com/design/intarch/
The fact is these are not widely publicised products you will find on their standard microprocessor price list. But they do compete here.
Commditization - this is why IBM got out of the desktop PC business for all intents and purposes. If you work in the industry, you'll see it first hand as product managers continue to try and use price to win markets instead of using R&D and product differentiation. Apple is a great example of a company that uses innovation. HP is re-discovering their roots and hence winning back share.
On the momentum issue. Yes...in theory you're right. Just because AMD loses momentum doesn't mean Intel is gaining it. But we know who won 1 point of share and who won 3 points and we'll see what happens in Q4. You have your convication and I have mine. I haven't declared victory but as I said earlier, I was reasonably close on how Q3 would turn out. I'll make my Q4 forecast in 30 days once I get a sense of the WW holiday SKU line ups for the various OEMs.
FYI - I do have a draft overall analysis of Q3 for AMD & Intel which I've been too busy to finish. Will get that out soon and we can chat again -:)
Ashenman:
1. On embedded - actually, not all processors are offered. Only certain SKUs if you click down for you can see for example that only the E6400 and T7400 are offered:
http://www.intel.com/design/intarch/core2duo/index.htm
That's why this is an off price list offering. Because these get customised.
2. On share - nobody lost 1 point. I said someone won 1 point (AMD) and somebody won 3 points (Intel) which is a very different trend from what we've seen over the last few quarters.
3. The entire commoditization discussion may be a little moot now seeing Dell's AMD notebook offerings are not cheap. You're right - let's see how it works out. I think HP will gain share and also lift margins with their new strategy as they reverse their direction which was really focussed on trying to use low prices to win share. Which didn't work out well for them because nobody can beat Dell at managing the supply chain. Not even UPS...I wonder if Dell is in the wrong business -:)
Where did you get the "Sub $500" figure that you display so prominently in the title? Even Dailytech refutes the "Sub $500" claim and gives the correct price of $549. Let's see what is really at Dell:
Inspirion 1501 with Sempron 3800+, $549
Inspirion B 1301 with Celeron M 380, $499
That's $50 difference. $10 of this is the LCD display. The other $40 would appear to be the slightly bigger battery and slightly faster processor. So, where is the calamity that you have been proclaiming?
BTW, I checked at Gateway and they don't offer any Merom notebooks, only Yonah. They also offer Turion, Sempron Mobile, and Celeron Mobile. Why no Meroms? Is this a capacity problem? Is this going to cause Intel to lose customers?
Yes, I am well aware that when the actual figures come out that my percentages could change slightly. The fact will still remain that Intel has not gained volume share or perhaps has gained only a tiny bit, not enough to matter. This statement But we know who won 1 point of share and who won 3 points is nonsense. The actual change in share will be less than half a percent. Essentially no real change since Q1. In other words, at the rate AMD has gained since Q1, it would take 25 years for AMD to gain enough share.
And, yes, it is true that AMD is not gaining as fast as it did in Q4 05 and Q1 06. However, it does appear that AMD has consolidated its gains and is not losing. I said after the Q2 earnings that the share gains could be just a bump and could fall back again. That hasn't happened.
So, where are you getting the idea that Dell is going out of business or driving AMD out of business? Also, I still have no idea where you get the notion that there is some kind of blood feud between Dell and Intel. Dell sells Pentium M, Yonah, and Core 2 Duo notebooks.
"Check out scientia's blog. He puts the actual marketshare aspect into perspective, concerning VIA. AMD is still growing, which is bad for Intel's current business image (though not necessarily for Intel overall)."
Scientia assumes both Intel and AMD ate Via's market share (which presumably is low end). The only issue with this is there is no way of knowing if this is true - it is just as possible that AMD ate all of Via's low end position and Intel took from server and higher end desktop. This is just another possibility - I'm not saying it is the case but Scientia's analysis is based on speculation. So if Via rebounds (I doubt it); they could take the share back from only Intel, only AMD, or a combination in between - there's no way of knowing... a couple more data points are needed (2-3 quarters).
He also postulates that Intel should have gained more share with Core2 people seem to forget these products have been out ~3 months and are still in relatively low volumes...Wait until mix gets greater than 50/50 (Q1'07)and then draw conclusions
MR 180
Intel is getting pistol-whipped by a nemesis 1/10th its size. Every AMD notebook Dell sells is one Intel can’t, even if worst case scenario of small margin is realized. It’s true Dell maybe eating the entire product line for now, but not for long. This is war remember, stop crying when you feel a little pain.
Scientia:
1. Here is the link to the story on sub 500$:
http://dailytech.com/article.aspx?newsid=4706
2. Please see my post carefully - I said death spiral of commoditization. That does not mean Dell is going out of business - I've confirmed that in the comments. The implication is their business model is is taking them into a space where their only strategic differentiator is cost and if you study business strategy, you will know that this is not a permanent sustainable position as is obvious from their share and revenue declines. I repeat sir - my overall observation is Dell is in a hole on where to go next strategically and they are struggling to find a new direction. As a result of which they continue to do what they've done in the past...focus on driving cost down.
To repeat my earlier observation, I thought they were using AMD on the low end for share and Intel on the high end for margin from the desktop pricing. But seeing the notebook pricing which actually makes reasonably comparable specced systems of AMD even more expensive than Intel I am not sure that generalization is accurate any more so ignore that for now.
Ashenman - the embedded business is highly customized to a customer's requirements - it's not just the cash register down at the supermarket. For example, you may have specially built tablet notebooks for army use, etc where the customization can happen down at the chip packaging leve. Think of it as a made to order business. Also, both Intel and Via (and AMD) offer a range of products in terms of performance, size, thermals, etc.
On the Walmart comment - you're probably right but with two exceptions:
1. Dell is in the technology business and people expect innovation. Cost will take you only so far.
2. Look at Walmart - even their astonishing growth has slowed down so I reiterate that using cost is not a permanent sustainable strategy. in fact Walmar has reached a point where they are now innovating - with things like their own shelf brands, setting up online movie distribution, etc.
Dell has not figured out where to go next and this is the issue.
I thought they were using AMD on the low end for share and Intel on the high end for margin from the desktop pricing.
They use both AMD Sempron and Intel Celeron at the low end; almost no difference. At the low end of the notebook market they use Sempron mobile and Celeron mobile; again, almost no difference. I'm really surprised that you would say that about using AMD at the low end for share. This isn't 2002 and 2003 when AMD sold K7's to compete with Intel Celerons.
I can understand why you no longer find the need to write your blog. The other side is just making a mockery of itself that it doesn't even deserved a mention...
Scientia - I was referring to their business/revenue strategy...not their SKU line up. I'm talking about using Athlon X2 for share and Core 2 for margins.
DO NOT POST
S180, I'd like it if you did not moderate comments for immediate posting:) Feel free to delete flames/spam or whatnot.
Post a Comment