Sunday, September 03, 2006

New Bed-fellows

With ATI and Dell no longer close partners for Intel, I'm betting that the impact will be to draw Nvidia and HP closer to Intel. Last year Intel signed an MOU with Nvidia for joint development. While I don't believe I saw details of what they would develop, I'm sure right now both those companies are discussing how to get close and comfy. As for HP, they are focussed on displacing Dell. At last they've remembered their roots and have started to innovate with their products. Their new marketing campaign (they hired Satjiv Chahil from Apple) does not seem to be hurting them either. Now that Dell has embraced AMD, HP will use this to gain leverage with Intel and slide deeper between the sheets with them. Specially since Intel now has some processors that do not suck.

While both these would be relationships of convenience, they would not bode well for AMD if they materialise. Some details of Bearlake are now available and dual GPU is supported. I'm inclined to believe Intel will support SLI over Crossfire. I'm sure there will be more coming from Intel & Nvidia.

On the HP front, I believe they will more and more leverage Intel's platform solutions and possibly brands to differentiate themselves from the others. Their renewed focus on innovation around the PC will sit nicely on top of the platforms. I think they will continue to gain share at Dell's expense. And ride higher ASPs as Dell's decline because they are bringing more value to the end user. My guess is 3 quarters from now HP will be more concerned about Apple (at least in mature markets) than Dell. We'll go there later. Dell will still be a screwdriver company searching for direction. And AMD will have won them at a point when their star is descending. Right now HP knows they can get any price they want from AMD since they had a common enemy - Dell+Intel. Now, they will always suspect AMD may be giving Dell better pricing and their unified stand against Dell+Intel is going to fray.

On the pricing front, Intel's uniform pricing for desktops is inevitably going to trickle down to notebooks. The desktop pricing change was probably to reduce the gap between the OEMs and the unbranded channel where Intel sees a higher ability to lead the market to their newest products. But there is also a huge upside on OEM relationships by taking the suspicion that a competitor is getting a much better price out of the equation. This may not help HP or Dell but it helps the smaller OEMs...specially local OEMs. It may even help the guys at Voodoo. Right now Intel is milking their lead on notebooks but it's only a matter of time before their smaller customers start demanding uniform pricing there too. Once this happens, AMD has a couple of paths. They can continue to use discounts to win deals. However, the flip side is their customers will be telling them Intel doesn't surprise me...or stab me by giving my competitor an even better price. The other route is to follow Intel and also establish uniform pricing. This I think will be good for the industry overall. The PC business is sadly filled with too many players not adding any innovation and instead have almost a commodity trader's mentality. Where price is everything. Bringing innovation back and reducing the focus on price as a key advantage is something I think can only be good for the industry. Can you imagine if we had another 3 companies who were able to innovate like Apple???

So here's where I think this is going. Intel will drive uniform pricing. AMD will not because they don't have the size or position yet to afford to walk as much as they may like to if their customers squeeze them on price. Intel will fight the price war as long as they need to till AMD starts to bleed. It's what they've done every time and they will do it again. ASPs will spiral down over the next 2 quarters as will profits. In Q4 AMD buys ATI for 5.4 billion. At which point they have borrowed 3 billion and just the interest on that will probably wipe out the quarterly profit they're making. Intel will drop prices again before the end of the year and bring Conroe mainstream. Merom is entering at exactly the same price points as Yonah so that transition is a no brainer. At which point AMD will be forced to drop prices again to ensure their factories are filled or gross margin goes through the floor. For the next 6-9 months there is blood everywhere. At the end of which AMD will have emerged in worse condition financially. Though Intel won't be sparkling but they fact they're cutting costs now before things get further out of hand will help. The other pressure tactic from Intel is the introduction of a new product almost every year and a new architecture every 2 years. At which point AMD will be forced to make some difficult choices on where to use their resources - protect the existing PC franchise or invest in building platforms. Hector will be forced to cancel or scale back some of his initiatives. Perhaps things like PIC will drop off. We'll still have a duopoly but AMD will risk being a player only in the PC category while the exponential growth comes from new initiatives like platforms, next generation wireless - Intel has a lead w/ Wimax, convergence of PC & CE (kind of linked to platforms) where Apple will rule and HP will wish they did, etc.

12 comments:

Joshua said...

lol Sharikou 180 wait for a year then start asvertising ur blog. I was kidding but make a bigger amount of articles so peeps have more to read!

180 Sharikou said...

-:) Joshua - I get your point. I'm not running around the Internet pumping my blog everywhere partly because it will take time to get some critical mass - articles and participants. I'm hoping more folks will come in and begin adding intelligence and analysis...and hopefully provide new trains of thought that we can explore. I'm not pretending to be an expert (am definitely not a Ph D). But everyone has to start somewhere.

cheers...

Anonymous said...

180:

If you want to start the ball of discussion rolling,
you need to kick the ball around and not just toss it out there.
For example on your blog on the AMD/ATI merger FUD, it is too generalized. You need to add more of your own opinions on what kind of risks AMD has not forseen.
Maybe Hector Ruiz will suffer a nervous breakdown or fist fights broke out in meetings?

180 Sharikou said...

I get your point. You're right...I need to kick off and wait for the players to line up their side.

Scientia from AMDZone said...

Your descriptions of HP are a bit strange. HP has been close to Intel for a long time; HP is the only company pushing Itanium which was a joint project between Intel and HP. However, HP has always used AMD chips even at a time when practically no one but eMachines and TigerDirect were using them.

The real shift in computers occurred a long time ago when Intel starting making and selling motherboards. Before Intel started making motherboards larger companies were at an advantage because smaller companies could not afford the R&D costs. That advatage is now gone. This took away the advantages that IBM, HP, and Compaq had. This is why Compaq was bought by HP and IBM sold its PC line. This is also why HP has been slipping.

You might recall as well that Packard Bell was once the dominant value system maker. PB was pushed out by eMachines. Gateway was once the dominant mail order company but they've been pushed aside by Dell. Gateway has now purchased eMachines to have a price point under Dell.

Dell and Gateway both were once Intel-only shops. Now, both are using AMD as well. HP and Gateway both resent the fact that the "Intel Inside" campaign benefitted Dell more than it did them. There is no indication however that either HP or Gateway are getting closer to Intel as both have increased their AMD offerings. Nor is there any indication that Dell is moving away from Intel. Dell simply found itself displaced by Apple as Intel's favorite and so it is no longer limiting itself to Intel. This was also shown in its purchase of Alienware which uses both Intel and AMD procesors.

To suggest that any of this represents a move toward either AMD or Intel is incorrect as even TigerDirect and eMachines sell Intel based computers. Since Intel sells 3.5X as many processors as AMD a company would have roughly 1/4 of its sales in AMD systems. However, since half of Intel's business is in Asia this figure could be higher in both North America and Europe. Probably, 1/3rd would be closer for companies that don't do large volumes in Asia.

The notion that Dell will simply decline and leave a vaccum for HP to fill is incorrect. Dell is currently looking to increase the value of its server line which brings them head to head with Sun, HP, and IBM. Dell's purchase of Alienware also shows that they are looking for higher value systems rather than just lower price as Alienware is definitely an upscale brand.

HP is not in a position to get any price they want from AMD. AMD dropped the low price model when Ruiz took over. I'm sure that HP is a good customer but there are also large customers like Google that you never hear anything about. Sales like this will only increase now that AMD has ATI.

Your estimates about Apple are overly optimistic. Apple no longer commands enough of a price point to make large profits on its hardware and Apple is shackled very tightly to software from MicroSoft. This is a complete reversal from the point when Microsoft started writing a little software for the Mac. Today, Apple cannot introduce a version of its OS for PC's as it would like to because MS would stop supporting their office software for the Mac. The logical merger (or at least partnership) for Apple would be Sun because Sun does large systems but not desktop systems while Apple does desktop systems but not large systems like Sun. This would take Apple out from under the MS threat since Sun already provides Solaris for X86 systems and owns Open Office software. As it is today, I don't see Apple's increasing its marketshare.

Your assumptions about the notebook market are also incorrect. AMD will not gain marketshare in notebooks by having a lower price (since this is what mobile Celeron and Sempron are for) but by using ATI to build a better mobile chipset. AMD will get a drop in power draw on Turion just by moving to 65nm at the begining of 2007.

I find it puzzling that your entire argument tends to revolve around marketing which has been Intel's advantage instead of engineering which has been AMD's. The reason why AMD does not have to have value pricing to compete is because today is nothing like 2002. AMD now has server chips and high value brands. AMD is picking up corporate clients and large volume customers. AMD is adding virtualization and RAS features that are better than what Intel is providing. Even MicroSoft has decided to support AMD's Pacifica virtualization. AMD now has 300mm production which it did not have in 2002. AMD will be at a good volume (50%) on FAB 36 by the end of the year. From that point on the cost of producing chips gets lower every quarter. Ths means that even if AMD charges less they will not actually be decreasing their margins and the reduced price will be made up in volume. By the end of 2007 AMD will be in better shape than they are today, not worse.

You've also overlooked AMD's introduction of native quad core in mid 2007 which will easily beat Intel's MCM version. Presler is MCM, Yonah and Conroe are native dual core; it makes a big difference. However, not only will K8l be released in quad core but dual core also which should put AMD back in the lead in terms of SSE but maybe still a little behind in Integer operations. AMD will also introduce a new mobile core in 2008 and will upgrade K8L from Direct Connect architecure 1.0 to 2.0.

Finally, Intel's plan of introducing a core refresh every two years is actually less sophisticated than what AMD is doing. AMD is changing the actual die layout to a modular configuration. This will allow them to refresh the die as often as every six months but more likely on a yearly basis. This will easily allow them to keep up with Intel's two year refreshes.

It is curious that a big part of your argument is based on price and the higher performance of C2D when only 8% of the processors sold by Intel in 2006 will be C2D. Even in the last quarter of the year, 77% of Intel's sales will be Prescott based Pentium D chips. AMD can easily compete with the Pentium D's. You proably are not aware that from the first quarter of 2006 to the last quarter of 2006 AMD will double the ratio of Athlon 64's to Semprons. AMD is moving its market segment up and leaving the Pentium D's to the lower value, Sempron segment. Intel will see negative growth from 2005 that is several times greater than its original 3% drop. I would estimate a 12% drop. In contrast, AMD should see growth of at least 20% versus 2005.

The above figures are counter to your base notions about AMD and Intel. You seem to associate success with Intel. Therefore if a company like Dell moves away from Intel they will decline while a company moving closer like HP will be successful. Intel's first and second quarters showed about 15% drop and Intel projects its third quarter to be even worse. At 23% of sales in Q4 there is no way that Intel can make up the decline even with C2D. And, companies clinging to Intel in 2006 would see the same drop so it makes sense to sell some AMD.

Sharikou, Ph. D. said...

In the x86 industry, whoever has the high end server market dominates. I have pointed out this before. There is a story in the AMD lawsuit: Compaq wanted to do a high end Athlon desktop (Athlon was the top desktop performer then), Intel heard of it and delayed supply Compaq Xeon chips. Compaq had to cancel the AMD product line.

When you can make $10K by selling one high end server, that's the profit of seling 200 PCes. Also, when you sell a high end server, you have a good chance of also selling a bunch of PCes to the company who bought that server. So whoever controls the high end server market is the dominant force. And that dominant force is now with 8P 16 core Opteron machines.

So, will HP and Intel get closer because DELL and AMD are now partners? They may, or may not. That depends on how their management view the situation, and their management may make the wrong decision. However, from now on, AMD's market share will mainly determined by customer choice, instead of monopoly dictatorship. That's all that matters.

In any case, Intel is finished. You think Intel will give Nvidia a bigger piece of chipset pie to hurt ATI(to be part of AMD)? No. Intel has a chipset business and a lot of Intel workers' paychecks depend on that business. Intel will have to reduce 3rd party chipset production , as its CPU market share is shrinking and it needs every piece of chipset to make a few dollars more.

http://sharikou.blogspot.com

"Mad Mod" Mike said...

The problem becomes reality when you step forward. On my blog and Sharikou's, we are not being general and not staying neutral, we give our 2cents on the subject and when we do, people bite down our necks for being biased.

180, you will be branded a fanboy as soon as you do it too, whether it's for AMD or Intel, doesn't matter. I suggest doing it anyway, you will attract alot of criticism (I used to get endless spams of 100's of messages, that ended thankfully) but you might get a few good responses out of it.

PS: I'll be adding this to my article section on my website (www.RubyWorks.net) if you don't mind.

180 Sharikou said...

Wow - that's a long comment from Scientia. I'm travelling right now and don't have the time to respond in detail. I will come back to it in a day or so. My one observation from a quick scan is I think a couple of my observations have been misunderstood.

1. I'm not talking about price driving the market - which is the case right now. Quite the reverse. I'm hypothesizing that we may be heading back to a place where OEM innovation is the driver.

My comments on uniform pricing are really about how they can change the dynamics of customer relationships. Specially smaller and more local OEMs.

2. HP's sales & marketing teams have resented the Intel Inside program. They do not like Intel (in their minds nothing more than a supplier) having so much leverage. They have been the primary driver of acceptance for the AMD brand with end users (not just corporates). It's not high, but it's what it is thanks to a lot of help from HP's marketing teams.

3. I'm also not suggesting Intel will abandon Dell or Dell will give up the market. However, HP will be more open to pushing Intel since their biggest competitor has adopted a second supplier. They have more reason to hate Intel less...or at least have a one night stand from a sales & marketing standpoint.

On a seperate note, I think their Itanium patience is wearing a little thin. If I remember correctly all the HP engineers working on Itanium were absorbed by Intel a few months ago.

4. On Apple, you're focussing on the enterprise space. I'm talking about consumer and specifically the convergence of PC and CE. They will provide the true convergence and the ability to enjoy your digital content around your home from your Apple Mac or Macbook. They will give you great hardware (based on Intel platforms), fantastic software (OS + apps), excellent usability (remote control) and all the content you need...movies and music. This is where they will grow...and I'm anticipating new exciting products that will begin to deliver this next year. I also believe their reliance on Microsoft should be reduced with Bootcamp and then possibly Virtualization built into Leopard.

5. Notebooks...I'm not suggesing price will drive AMD share. Again, my observation is Intel will at some point move to a uniform pricing scheme there too.

6. Yes - my arguments are based on sales & marketing. Because if business were won merely by having the best technology then AMD would have almost 100% market share in servers & desktops. But that's not how business works.

7. I'm not sure where you took out my argument is based on price and C2D performance. I'm talking about uniform pricing where the price list is the price everyone gets. And I don't believe I even mention C2D let alone refer performance.

8. On the rate of refresh. Processor design is only one part of what it takes to bring to market a new product or even a new derivative of an existing product. There are multiple business issues and costs associated which is where size and scale on multiple aspects come into play.

9. I completely disagree with your premise on AMD having high value brands. AMD's brand is high value to large corporations and people like us. But walk into a PC mall in an emerging market or ask an SMB and they don't even know who AMD is. Their brand is in better shape than it was 5 years ago but it's not high value.

10. Your premise that Sempron is fighting Pentium D is wrong - they are not even in the same price band. AMD was living in the seam (~90-100$ price point) between Intel's Celeron (50-80$) and Pentium (120$ +). That seam has now been closed with Pentium 4 and Pentium D as of July. I am convinced Intel will kill off Celeron over time and replace all their low end with Pentium.

11. I'm not going to argue your point on K8L performance, etc. There's no data but more importantly I'm not concerned with that aspect on this post. As I said before, having the best technology doesn't = success in the marketplace. I'm focussing on the business issues here. I'll also be the first to admit that I don't have the technical smarts to win an argument with you on that -:)

12. On the refresh cycles - this is where I wish I had the technical know how. What you're describing for AMD sounds simlar to a couple of articles I read on mini core. But I won't pretend I understand this completely. However, I do think that this business will become like the car industry. Highly segmented with different products excelling in different areas.

13. I think it might be worth reading my last para again. Yes, I think Intel will succeed in the longer term. But not because they will necessarily have the best technology for the traditional PC business as we know it. They will succeed if they can reasonably protect their current franchise but more importantly ride the next waves of growth through their platforms. And I mean more than just the basic ability to design and manufacture CPU + chipset. AMD could not have pulled of Centrino - they just don't have the scale at this point required to move an eco-system or even fund and verify 300k wi-fi hotspots worldwide for Centrino.

Whew - I actually did respond a lot more than I thought I would. I'm hoping other's on the blog will be able to jump in and analyse the technical related points to the level they deserve. As I said, I need to pick sides on each of these arguments to provide the stimulus for discussion and it seems to be working.

Scientia - thanks for the long note.

Scientia from AMDZone said...

My comments on uniform pricing are really about how they can change the dynamics of customer relationships. Specially smaller and more local OEMs.

I'm hypothesizing that we may be heading back to a place where OEM innovation is the driver.


Okay, your argument is not about the price of the chip itself but price to OEM's. You are theorizing that uniform pricing will allow more favorable terms with smaller OEM's and that these will then drive innovation.

I would have to disagree. Smaller EOM's like VoodooPC and Alienware only account for a tiny fraction of the market. This isn't where AMD and Intel make their money. At any rate, there is no reason why this would favor Intel.

3. However, HP will be more open to pushing Intel since their biggest competitor has adopted a second supplier.

This doesn't follow. Dell, Sun, HP, and IBM are all competitors in the server market. Yet, all four increased their AMD offerings earlier this year. By your way of thinking some of them should have decreased their AMD offerings to be more different from their competitors. Also, it makes no sense for HP to offer fewer AMD systems since this would make them more like Gateway and Gateway isn't doing as well.

On a seperate note, I think their Itanium patience is wearing a little thin.

No. HP's Itanium sales were very good last year and had doubled from the previous year. I'm certain they will keep pushing Itanium for the next several years.

4. On Apple, I'm talking about consumer and specifically the convergence of PC and CE. They will provide the true convergence and the ability to enjoy your digital content around your home from your Apple Mac or Macbook. They will give you great hardware (based on Intel platforms), fantastic software (OS + apps), excellent usability (remote control) and all the content you need...movies and music. This is where they will grow...and I'm anticipating new exciting products that will begin to deliver this next year.

Everyone is doing this. However, there is no indication that this will drive the PC market. The indications are that it won't. However, looking at your descriptions:

true convergence
great hardware
fantastic software
excellent usability
new exciting products


It is clear that you are not very objective in your view of Apple.

6. Yes - my arguments are based on sales & marketing. Because if business were won merely by having the best technology then AMD would have almost 100% market share in servers & desktops.

Well, you are leaving out two things: First of all, Intel has had marketing budget of $8 Billion a year for the past several years and this has not stopped AMD from gaining marketshare. Secondly, the reason why your 100% marke share conclusion for AMD is invalid is because AMD has not had enough capacity to serve that much of the market. Intel has had multiple 300mm FABs for years. AMD didn't have a single 300mm FAB until 2006 and it will only be at half capacity by the end of the year. AMD's market share has grown as its capacity has grown in spite of Intel's marketing. Marketing is not the deciding factor.

8. On the rate of refresh. Processor design is only one part of what it takes to bring to market a new product or even a new derivative of an existing product. There are multiple business issues and costs associated which is where size and scale on multiple aspects come into play.

AMD produced 45 Million chips in 2005 and should do about 65 Million in 2006. In 2005, AMD produced more chips than IBM. AMD is large enough that this isn't a factor.

9. I completely disagree with your premise on AMD having high value brands.

No. AMD's brand value is higher than it used to be in, say, 2002. AMD's average selling price has been increasing but it is still below Intel's.

10. Your premise that Sempron is fighting Pentium D is wrong - they are not even in the same price band.

I may not have phrased this well. Basically, what I was referring to is that the Prescott based P4's have decreased in value (mostly due to C2D). However, a dual core Pentium D is obviously worth more than a single core Sempron. As these prices push down they squeeze the low end and make it less profitable so AMD will make fewer low end chips.

AMD was living in the seam (~90-100$ price point) between Intel's Celeron (50-80$) and Pentium (120$ +). That seam has now been closed with Pentium 4 and Pentium D as of July.

I have no idea where you got this notion from. AMD's chip volume is about 30% of Intel's chip volume. This doesn't fit into a seam. In reality, AMD's Sempron's compete directly with Intel's Celerons in the same price range. Also, AMD's single and dual core chips compete directly with Intel's in the price range above that. I can assure you that AMD does not manufacture the majority of its chips to sell in the $80-120 price range. Also, while Intel's Celeron price range tops out at $70, AMD's Sempron price range tops out at $100. The X2 chips which account for 3/7ths of the Athlon 64 chips range from $150 - $300. The price range you mentioned would be about right for the current single core Athlon 64's which account for about 1/3rd of the chips AMD made for the 3rd quarter.

13. Yes, I think Intel will succeed in the longer term. They will succeed if they can reasonably protect their current franchise but more importantly ride the next waves of growth through their platforms.

In other words, Intel will succeed if they succeed?

AMD could not have pulled of Centrino - they just don't have the scale at this point required to move an eco-system or even fund and verify 300k wi-fi hotspots worldwide for Centrino.

Your assumptions here are incorrect. Wi-Fi involved networking giant Cisco and its subsidiary Linksys. Intel only invested $150 Million in Wi-Fi. Yes, it does have a Centrino hotspot verification program but this is trivial and could have been duplicated by AMD for no more than a couple of $Million.

In fact, your assumptions here are very puzzling. There are several directories of campsites in the US that are quite thick and required much more effort to research and compile than simple wireless access. These directories were created by companies much smaller than Intel and are updated every year.

You are correct however that when the Centrino campaign started AMD did not have enough mobile volume to have had the same effect. AMD has grown tremendously since early 2003 however and is reaching that point.

180 Sharikou said...

Scientia... -:)

1. To be clearer about my comment on smaller OEMs - it's not just about the terms they get, but about how uniform pricing affects the trust factor. Uniform pricing helps reduce the focus of a product manager at a Tier 2 OEM wondering how to hit his qtrly number on just price because they're always thinking “the next guy is getting a better deal than me...and Dell + HP definitely are so I need to drive my price down to compete”. Once you know everyone is getting the same price, you start wondering how else you can compete because price is removed as a variable. Now there are other component mfrs they negotiate with but the Intel cost inside the BOM (bill of materials) is high.

You're right, my comment on Voodoo was flippant and they shouldn't be compared to a Tier 2 OEM in terms of size & scale My job exposes me considerably to Asia (I'm writing from China today) and the scale of a Tier 2 OEM in a place like China is significantly larger than a Voodoo or Alienware. We're talking millions of units annually sold to the mass market in what is the fastest growing market in the world. Folks like Voodoo are niche and small players. The dynamics in a market like this are mind-bogglingly different. 15 years ago, Lenovo was Legend...a small importer and assembler. There are multiple companies out here with similar ambitions growing nipping at Lenovo's heels on their home turf and stealing market share.

I wasn't really suggesting that uniform pricing would favour Intel, just that AMD would have to decide whether to follow a similar path or not and how that could play out. Having said that, Intel does have a couple of advantages for Tier 2/3 OEMs and the DIY channel which is reach of sales force and that all these guys want to leverage the Intel brand because they don't have their own brands to speak of. Of course they also want to leverage Microsoft, ATI, AMD & Nvidia's brands to varying degrees but from a brand perspective MS and Intel are the ones they try to get access to the most. AMD's advantage is a growing acceptability and till now lower price.

3. HP push for Intel - SKUs do not = market push. HP has been pushing AMD aggressively for many reasons (while taking sweet pricing deals and marketing $s from Intel). One of them is to keep a check and balance on the Intel + Dell relationship. Perhaps another is they would love to see Intel be cut down to size. With Core 2, Intel finally has a competitive product. To sustain this check & balance now that Dell has gone AMD, they will use Intel. I'm not suggesting less AMD, I'm suggesting more go to market push for their Intel SKUs. I also said one night stand - it lasts as long as it's convenient to HP.

On Itanium, you're right on volumes. I think what I was trying to say is they reduced their cost and energy exposure to Itanium development. They are no longer willing to throw endless money here while they do want to reap the revenue they can. But that’s my perception

4. Apple - you and I will have to disagree on who's POV is objective. Nobody is doing this convergence well. Everyone is talking about the Digital Home but nobody has truly delivered it. So far Apple has been extremely successful in designing consumer products that consumers want to buy. They have the entire chain - from hardware to content. I believe they will succeed and much faster. The fact that they are growing off a small base should not throw us - I reiterate they will grow and this aspect will start to worry HP if it isn’t already.

While your observation that there's no evidence this will drive the PC market is technically correct - there cannot be any merely because the evidence will only exist once it happens. There was no evidence that the iPod would drive the personal audio marker but hindsight is always 20/20 so let's wait and see. I'm hypothesizing and only time will tell.

6. Well...my comment on 100% was really an exaggeration to make a point that it's not just the technology. Having said that, I'm not changing my POV. Just having the best technology does not decide the winner in the marketplace. If AMD could theoretically supply 100% of the marketplace I do not believe they would have 100% MSS even if they had the best product. Same holds true for Intel. That's my perspective...agreeing is your call.

I’m not sure what the source of the 8 billion $ number is. Intel’s 2005 annual report shows “Marketing, general and administrative costs at 5.6 billion”. Even their largest marketing program - Intel Inside is about 1.5 – 2 billion so 8 billion sounds a bit extreme. I'd say if you add it all up (w/o h/c) their marketing spend might be 2.5-3 billion. Again, I’m thinking traditional marketing so it might be a difference in definition. However, to spend almost as much on marketing as they do on R&D and fabs combined sounds extreme and I don’t see a basis for that number in their annual report.

8. I wasn't talking manufacturing scale here. I was talking about “go to market” scale. There is a huge amount of energy and money required to bring a new product, architecture or brand to market and gain momentum for it globally. To do this at an annual frequency requires significant money and resource. It’s not that AMD can’t do it – but even after losing 10k jobs, Intel has more resource and money. My point was this will add to the choices AMD will need to make on how they use their resources.

9. We’re both in agreement – AMD’s brand value is increasing…perhaps faster in the corporate space than consumer. But it is still significantly lower than Intel’s. Business week publishes a report every year with Interbrand on the best global brands here and the Intel brand (assume they were to sell the Intel brand as an asset) is valued at 32 billion $.

10. We’re in agreement - I’d conjecture that over time Celeron and Sempron will lose sales focus for both Intel and AMD as nobody makes money at selling such low priced products for long periods of time.
On the seam – I am talking about a pricing seam in Intel’s old price strategy. Lenovo who have been a huge customer for AMD have split the China market open over the last 2-3 years with Athlon 64 that they have been getting at 90-100$. Intel was non competitive because all they were throwing was crappy Celeron at 70 bucks or then Pentiums which Lenovo wanted at 120$. This is the seam I’m talking about. The price we all see on Newegg are not the prices Intel or AMD offer Dell or Lenovo or will take in to win a large tender or deal. Please also consider that I am talking about what’s been happening over the last 2 years as AMD has made tremendous headway. I think in a couple of qtrs we will have a better idea of what’s happening to both co’s product mix as a result of the price war.

13. Intel will succeed if they succeed - well no…that’s not what I’m saying. I’m saying Intel will succeed only if they can effectively bring their platforms to market the way they did with Centrino. So far, Viiv appears to be a disaster to me, ultra mobile PCs (UMPC) have not got off to a good start, time will tell on VPro and Wimax is still far off and their Community PC still seems to be at childbirth. I’m saying to grow, Intel must execute well on these platform initiatives and so far I haven’t seen much evidence of them doing that.

AMD pulling off Centrino – I’m not sure what you’re reference to camp sites has to do with my comment. I don’t think you understand the resources and scale involved to go cut deals with ISPs, retailers, coffee shops, airports, etc across the world to set up wi-fi hotspots, verify them for Centrino and then brand them. You have got to be kidding me if you think this can be done with a couple of million $s. Even today AMD still does not have the resources or the money to create a Centrino brand equivalent. Let me put this in context – Intel’s Centrino revenue is larger than all of AMD’s revenue in 2005. So I repeat they do not have the scale as a business. A quarter from now AMD intends to introduce their first mobile “platform”. Let us see whether they can replicate or surpass Centrino in terms of Centrino as a business and a brand.

Anonymous said...

If Intel is to cut off ATI totally from its product lines and relies solely on Nvidia and Intel's own graphics products, Intel is making a big mistake.
Why?
Because we've seen one product from one company will outperform another's in different ranges at different times, so Intel is making the same mistake as Dell, namely no diversity to hedge the dynamics of the industry.
The graphics card is just as important as the cpu in gaming, CAD design and other imaging intense works.

180 Sharikou said...

I agree. Which is why I think Intel is now scambling to increase their investment in their graphics business and the speculation is they will introduce discrete graphics middle next year. See my reaction:

Link