Been busy this weekend and just able to grab a little time to listen to the earnings properly:
1. Server revenue, double digit unit growth and ASP growth - this is a very healthy sign. As I had said earlier, Intel is starting to win back their big losses in servers...specially in the DP space. Combine that with the fact that Woodcrest is already 40% of the server market by volume in 3 months (as announced at IDF first). However, the impact of this may be felt less to AMD if they decide to retain premium pricing on Opteron and really focus on the MP server market where margins are higher.
2. Mobile too had record shipments - at the cost of desktops which continue do decline as a share of overall market. Not unexpected as Intel does have a significant lead over AMD in mobile. Will be interesting to see the impact of Dell in this space in Q4. But with Merom ramping incredibly fast considering it would be the bulk of the volume in the Core 2 6 million units shipped, feels like they're in good shape here.
3. Inventory in microprocessors down QoQ. Chipsets and flash were up. Overall inventory was ~4.45 billion...a gain of 120 million over Q2. They also took a 100 million $ write off. But most importantly, all the WIP is 65nm. Which means the transition is happening quite rapidly.
4. Q4 outlook - revenue between 9.1 to 9.7 billion. Gross margin around 50%. On track to hit 95,000 heads from 102,500. About 500 million savings on capital spending and R&D against the original plan. MG&A costs to be about flat QoQ even as revenue increases. Overall, it looks like the effects of the re-structure are kicking in to the upside. The worrying thing here is the gross margin staying at 50%. Which means either the price war will continue or they will get hit with the 65nm and 45nm transition costs. Since the indication from Bryant is they expect to see prices firm n Q4, this is probably the former.
On the face of it, the qtr came through pretty much as I expected. Otellini believes they gained overall share but as I said earlier, we'll have to wait for Gartner/IDC to confirm. The most critical piece in here is servers and the Woodcrest ramp. The more share they re-gain here, the harder they hit AMD's margins and hence free cash which is very important as AMD takes incremental debt to fund the ATI acquisition. The price war should stabilize in Q4. Obviously Intel had planned for this when they forecast their margin at 49%. Overall, it feels like Intel is holding in mobile, winning back in servers and losing in desktop. While it's important to win back the desktop market just because of the sheer size, I think we will see the true impact of the price drops and the Core 2 introduction in Q4 due to the lack of lower cost motherboards.
My nett take away is the qtr went as planned. Woodcrest is the bright spot and some marginal upside in terms of revenue but nothing to write home about. A concern on margins for Q4 and the discussion on share still up in the air as Otellini claims they gained overall share and it isn't clear from a quick glance at AMD's call whether they believe they gained any. However, I'll be reviewing that in detail next and will see if we can call it. For now, Intel needs to stay the course and focus on ramping Core 2 desktop, get the right chipsets/boards into the channels and help their OEM customers get off to a quick start as they introduce products this quarter for Christmas.
I'll be reviewing AMD's results next and will be able to figure out what I think Q4 will look like based on that.
Subscribe to:
Post Comments (Atom)
10 comments:
Ashenman:
1. I'm not an accountant but I think the 100 million is over and above. What this means is they wrote off inventory they did not think they could sell. Which means they are growing inventory of what they think they can sell.
2. On servers, I think what we see happening is a re-capture of share in the 2P market for Intel. Since these swings take time, this is a surprisingly fast ramp for Woodcrest. I suspect even Intel is surprised. This should be AMD's #1 concern because as you say, the margins on the server business allow them to fund other things. If they get hit here then managing cash as they head to an ATI acquisition will be a potential issue. As I said earlier, AMD will have a tough time balancing margin vs share but will focus on share to keep the factories loaded. It looks like they have done exactly that but under continuous pressure this is not sustainable for them.
3. I think the inventory build up around chipsets may have to do with their issues on the Broadwater ramp. Also, there is a lack of lower cost motherboards out there right now for Core 2. Until that happens in Q4, newer chipsets will not necessarily move out that fast. On flash, I just think demand was lower than supply and they need to win more business. To answer your question: 37% was finished goods, 50% was WIP and 12% was raw materials. Roughly this is consistent with Q2 except they were holding more microprocessors at that point. Overall, a slight increase in inventory is not too bad considering it's chipsets. In fact, with their previous history on chipsets and ATI not being a supply source any more, it's probably a good thing to have some excess chipset inventory. Normally I would assume you would not move these automatically to packaging because you don't want to increase the cost of the inventory without reason. Doing it just in time optimizes the supply chain.
I actually don't think Core 2 mobile is a problem for them at all. Merom was a drop in product with price parity to Napa.
Ashenman:
On the inventory...you're probably right. My take is Intel's supply chain management is not as sophisticated when compared to someone like Dell so JIT is not 100% JIT. Also, I think they were forced to build chipset inventory for Q4 knowing ATI will not be supporting them. But I don't have a better answer than that.
On the mobile - my observation is Merom is a drop in at price parity to Napa so the high prices are probably what OEMs think they can get away with for a new technology. Not necessarily a premium Intel is making.
On the Vista comment - the industry is not expecting Microsoft's coupon to upgrade to Vista to really help Christmas sales so that may be less of a consideration for them right now and hence the higher prices. Also...you want to get rid of your old inventory before dropping prices on the new stuff.
On AMD, I'm not expecting them to screw up overall. But the one place they are making a hard choice is to support OEMs over their channel. Both in terms of supply and pricing. It's obvious that even with shortages in channel, they built inventories by 15% which means they either completely mis-called what their customers need or are being forced to hold inventory for Dell and other OEMs. Both are bad for channel. Hence, I think your confidence in system integrators being loyal is where they will come short and it's actually the OEMs who will be their friend.
On servers, as long as Intel does not have a competitive response to 4P, Opteron rules the margin side of the business so AMD has not much to worry in that niche. While the 2P stuff is volume, it's still better margin than client and I'm sure they won't give it up willingly.
I'm goin to look at AMD's Q3 shortly so I'll have a better perspective at that time.
Ashenman - AMD is screwing the channel that helped them get to where they are today. First, by not giving them supply and holding it for their OEM customers. Second, on pricing. Pls go back and read the earlier posts and you'll understand the issue and the fall out for AMD. It's a lack of faith. If I can't get product from my supplier and have to lose a sale, I'm going to start pushing the competition. Read this post:
http://sharikou180.blogspot.com/2006/10/dell-sucking-amds-capacity-dry.html
On Vista, interestingly Otellini was asked about Vista impact during the earnings call - you can read his take here but he does not believe this will move the market much to start with...at least in enterprise. I'm not sure what you mean by Intel introducing a Vista platform. Their 965 chipsets are already Vista ready for desktop. If you're referring to mobile - there are already plenty of notebooks based on Centrino that are Vista ready and Santa Rosa will add further graphics capabilities along with Robson technology. All you need to do is head to the Windows Vista site.
Well - all you gotta do is head over to the Intel web site and check out their desktop chipset product pages for their 965 chipsets which clearly state they are Vista Premium ready. Then mosey over to the Windows Vista site and check out the notebooks avaialble today ready for Vista...plenty of Centrino's down there.
On the OEM vs SI issue...both are important for different reasons. AMD in their focus on winning the high margin corporate business is making a similar choice. You gotta win and keep every customer. It is not a good strategy to decide to ignore either of your customer bases. If you're looking for precedent, Intel neglected the channel 2 years ago and paid for it. This time round, channel received Core 2 supply first - OEMs apart from Dell are still bringing SKUs to market because they are fundamentally slow to shift. Channel drives the momentum behind new technology due to their agility and the fact they carry almost no inventories.
I think we need to align on definitions first:
SI's are part of the channel = folks who sell unbranded PCs & servers. The difference between a SI (System Integrator) and a regular unbranded box pusher is the SI will try and offer other services like networking, tech support, storage, etc thereby trying to go up the value chain.
However, they both have the same problem when supply becomes an issue. Loss of customer = loss of faith in their supplier.
So I'm not sure what you're talking about when you say SI? Think it's better if you clarify too before we continue this discussion.
WRT Vista - Intel's existing integrated graphics are Vista capable. Go to the sites I indicated to confirm for yourself. The reason I believe the industry does not think there will be a huge instantaneous upgrade cycle with Vista this time is due to the corporate market not being enamoured of the product to accelerate purchases. Time will tell if they're right. Remember - AMD does not have a corporate platform right now. A platform is a set of hardware and software which is validated and certified to deliver a specific end user benefit. A platform is not merely sticking an AMD chip and ATI GPU together. AMD has said they will probably have this in late 2007 or 2008. There is a lot of work involved - more than R&D it's the validation that takes time.
Bottomline - do not expect either AMD or Intel to gain strategic advantage due to the Vista launch. The impact will be shared and equal - but probably marginal.
the qtr came through pretty much as I expected.
Yes, the only surprise to me was the margin drop for AMD. Apparently, the price cuts hadn't quite taken effect yet in Q2.
Otellini believes they gained overall share
and is deluded. Intel neither gained in terms of revenue, not volume.
but as I said earlier, we'll have to wait for Gartner/IDC to confirm.
We only need those numbers to determine to a tenth of a percent. But we already know that any change was tiny.
The most critical piece in here is servers and the Woodcrest ramp. The more share they re-gain here, the harder they hit AMD's margins
Incorrect. If Intel gains more it will affect AMD's revenue share, not margins. AMD's margins will actually increase in Q4 even if AMD were to lose revenue share.
Overall, it feels like Intel is holding in mobile
No, Intel lost some in mobile. Actually, the thing that you may not be aware of is that AMD's assault on mobile has not started yet. Intel isn't prepared for this.
winning back in servers and losing in desktop.
No. Intel gained in servers and also gained on the desktop. Servers are the real problem because AMD doesn't really have a change in 2-way until Q2 07. However, AMD does have a plan of attack for desktop so I don't really expect Intel's gains here to continue.
we will see the true impact of the price drops and the Core 2 introduction in Q4 due to the lack of lower cost motherboards.
I would guess that the true impact will be less than you expect. Intel's best position in desktop is Q2 07.
My nett take away is the qtr went as planned.
Meaning no actual gain but Intel stopped losing share.
Woodcrest is the bright spot and some marginal upside in terms of revenue but nothing to write home about.
Actually, AMD gained a small amount in revenue share. However, overall, it was a slight loss for AMD because of the margin decrease. In other words, AMD gained more money year on year in Q2 than Q3. The difference wasn't big but again not as good as it would have been without the margin decrease.
A concern on margins for Q4 and the discussion on share still up in the air as Otellini claims they gained overall share
Otellini is wrong. Intel gained no share.
and it isn't clear from a quick glance at AMD's call whether they believe they gained any.
If AMD gained any volume share it was a fraction of a percent. Not worth worrying about. Revenue share was the same, a fraction of a percent gain for AMD.
This should be AMD's #1 concern
Servers are AMD's #1 concern.
As I said earlier, AMD will have a tough time balancing margin vs share but will focus on share to keep the factories loaded. It looks like they have done exactly that but under continuous pressure this is not sustainable for them.
Your assessment is incorrect. AMD's margin will rebound 2% in Q4 unless Intel cuts prices even further. AMD's margin will continue to increase in 2007 as FAB 36 ramps. Intel does get some cost reduction from ramping but only about half of what AMD gets. AMD's margin will be back up to what they consider normal in Q4 and should increase again in Q1 and Q2.
On flash, I just think demand was lower than supply and they need to win more business.
Intel only lost $112 Million in Flash; that is better than they have done. Of course, if they were smart they would divest Flash.
on chipsets and ATI not being a supply source any more, it's probably a good thing to have some excess chipset inventory.
No, not if these aren't the newest chipsets. BTW, the reduction in processor inventory explains why Intel didn't gain in revenue share.
Normally I would assume you would not move these automatically to packaging because you don't want to increase the cost of the inventory without reason. Doing it just in time optimizes the supply chain.
Incorrect, JIT means less raw materials, not more.
I actually don't think Core 2 mobile is a problem for them at all. Merom was a drop in product with price parity to Napa.
Your assumption here is wrong. Intel will have an easier time with servers than mobile. As I've already said, AMD has not yet started its assault on mobile. Intel is not prepared for this.
It's obvious that even with shortages in channel, they built inventories by 15% which means they either completely mis-called what their customers need or are being forced to hold inventory for Dell and other OEMs.
Incorrect. AMD always builds up inventory in Q3 for Q4; this is normal. This inventory will drop to about 40% in Q4 if things go as planned.
Remember - AMD does not have a corporate platform right now.
You might want to rethink this. I'll quote from AMD:
Beginning on September 1, 2005, commercial, government, and education IT organizations will have a new choice in selecting PCs for their business needs. Organizations will have the ability to leverage the proven advantages of the AMD64 platform – the award winning platform for running today and tomorrow’s software – with great confidence through the AMD Commercial Stable Image Platform (CSIP)program.
Scientia - I'm wondering how to respond to your barrage of comments because you've missed the point in several places. I'd ask you to read:
1. Transcripts for AMD & Intel earnings calls along with analyst Q&A.
2. Mercury Research information on market share gain for AMD & Intel in Q3.
3. My posts including all the commentary relevant to this.
I repeat once again - just because you say it is so doesn't make it so. Bring the data and facts or just say it's your POV. However, get off the high horse dude and chill a little. I've said it before and I say it again. I am trying to call the next few months and I believe they will swing to Intel. When I believe it's swinging back to AMD...I'll call it like that. But at the right time.
Post a Comment