Still playing with the slice-and-dicer and thought INTC would be a good blue chip candidate.My rough cut EVA valuation gives with M* at $27. No real heavy lifting -- assume slow growth in capital base ~2% and with a healthy ROIC/WACC spread of 8%. Q3'12 IC was ~$62.5 billion (w/ no adjustments regarding cash).INTC trades historically at 2X its invested capital based (TA-NIBCLS). Using that metric on Q3'12 b/s implies a value of $24.40 (Firm) and $25 (Equity). Taking the 5 year invested capital average of $50.6B, one would get $20.37 -- which is roughly today's trading price.Despite upcoming tax implications -- the dividend yield is ~4.3%.So regardless of the noise around INTC (Apple shifting suppliers) and the market (fiscal cliff/dividend tax rate changes) -- I think a fair conclusion is there is some value here. Obviously it can get cheaper but I would think we are at levels that are justifiable on a more normalized business cycle time frame?Matthew(long INTC)
So regardless of the noise around INTC (Apple shifting suppliers) and the market (fiscal cliff/dividend tax rate changes) -- I think a fair conclusion is there is some value here. Obviously it can get cheaper but I would think we are at levels that are justifiable on a more normalized business cycle time frame? - MatthewI agree. I've been buying, lately, and will buy more now that INTC sells at <$20. I consider it practically a "no-brainer" decision.
It is only "no brainer" if you assume that Intel will continue to dominate the CPU market as they have historically. In any gorilla troop, the silver back is the absolute top ... until he isn't. In industry, a gorilla needs to keep using it gorillaness to keep winning the market in new generations and markets. Seems to me that Intel has allowed someone else to capture most of the cellphone CPU market (e.g., QCOM and ARMH) and they are seeing major threats in tablets. With Windows RT on ARMH, there is even potential erosion in servers.
Seems to me that Intel has allowed someone else to capture most of the cellphone CPU market (e.g., QCOM and ARMH) and they are seeing major threats in tablets. With Windows RT on ARMH, there is even potential erosion in servers. - tamhasI've posted my thoughts regarding Intel several times on several different boards, including several here on the Hounds. I agree that Intel has ceded dominance to others in the cellphone market, but I do not consider that an irrevocable loss. ARMH had great success designing low-power/low computational capacity chips for cell phones. Good for them! A whole lotta people on this planet still rely on 2G service. But I see the Smartphone/Tablet market evolving towards ever greater needs for low power/high computational capacity chips. Intel is a full generation ahead of competitors in developing the necessary chip architecture for such applications.QCOM is my largest holding in the chip sector, so I don't disagree with you there.Intel continues to be dominant in the enterprise server market. I'd be surprised if they lose significant market share in that niche. Big Dawgs continue to be Big Dawgs until a bigger Dawg comes along. Ain't seen that hound...yet.As for PC's? They ain't dead. Smartphones are cool and fun and, gosh, they can even be used as telephones! BUT, I don't do any serious work on a smartphone or a tablet (don't own either, actually). I still rely on a personal computer to earn my daily bread. So do/will a whole lotta other people here, there and everywhere (and AMD is dying before our very eyes).Intel below $20/share, 8.7 ttm P/E with a 4.3% dividend yield? Yeppers, I consider that a no-brainer (although, to be fair, I've been thinking about all this for three decades running).
I'm a bit of a fossil myself, since I do most of my work on a desktop, but even I am typing this on a tablet. There is a blurring of this space ... not the least of which being the move of applications to the cloud, making the client less relevant. The point is, Intel doesn't have nearly the lock they once had.
There is a blurring of this space ... not the least of which being the move of applications to the cloud, making the client less relevant. The point is, Intel doesn't have nearly the lock they once had. - tamhasTrue enough: ever more applications are moving into the "cloud" where they are hosted on arrays of enterprise servers. That's a significant reason why I like Intel's future prospects. Intel, as I've noted in the past, is found in roughly 80% of the enterprise servers that constitute the "cloud."
This might be a bit off beat, but the next tech wave will be from nano tech materials.......and nano tech chips of different dimensions and materials.......just some food for thought.....what is more serious about an investment in anything right now is that the market wants to fall.....if E waves are worth a study at all.....this is only the first wave down......there would be a couple of other waves to come....and this first wave might be far from over....corporate profits in many cases are falling......not good....then there is the 'fact' that could be.....five years or so till we get towards another bull market.....dont hold your breath buying intel towards what might be a top.....while the market might not get hot again for a good five years or more......buy the bottom......not the top......we clearly have not see even much of an over sold bounce yet......Dave
One of the biggest issues for server farms is low power consumption. With Windows now on ARMH, is Intel going to stay dominant in this space?
With Windows now on ARMH, is Intel going to stay dominant in this space?It is only Windows RT on Microsoft tablets that is on ARMH. The rest is still on Intel architecture.sf
One of the biggest issues for server farms is low power consumption. With Windows now on ARMH, is Intel going to stay dominant in this space? - tamhasThis is where my earlier comment about Intel being a generation ahead of the competition with its 3D Tri-Gate chip technology that offers robust computational capability coupled with low power consumption. Here's an article you may find interesting:http://tinyurl.com/anbznynChip manufacturing race between Intel, ARM tighteningThe race to manufacture the most power-efficient and fastest chips is gaining momentum, with contract chip manufacturer GlobalFoundries on Thursday announcing technology advances that analysts said could allow the company to catch up with Intel's chip-making capabilities by 2014.As matters stand, Intel's dominance in the enterprise server market remains solid for the time being. Coupla years from now? Who knows? It'll be a fiercely contested space.I'm comfortable owning INTC now and will continue watching the technology develop.
Yes, only RT now, but the big issue with RT is that there are only some applications for it. That doesn't make it ineligible for running specific cloud applications. I am not suggesting that Intel is going to disappear overnight by any means, just suggesting that they are not quite the silverback in sole domination of their space that they once were and that one should bear that in mind for long term investing. Short term, it is unlikely that we will get big, sudden surprises. But, don't close your eyes and go to sleep thinking it is completely safe because the landscape might change since it is changing already.
I agree. I've been buying, lately, and will buy more now that INTC sells at <$20. I consider it practically a "no-brainer" decision.The stock market overall is very highly valued from a historical perspective and will almost certainly experience a 25+% drop within the next few years (possibly much more).Intel will drop with the market. I see no reason to be long now.
A few comments on this thread, and your link. More of an omnibus reply.GlobalFoundries on Thursday announcing technology advances that analysts said could allow the company to catch up with This has been said by both GF and IBM for the past ten years or so, and all that happens is they fall further behind... but it will be different this time:-)Regarding servers, and windows RT.Most servers run open source (linux or derivative) so it isn't really a software compatability issue. It is absolutely a power issue, and one of the reasons Intel has all but shut AMD out of this market. For the lower power microserver market that ARM might be able to fit into, Intel has just announced the centerton CPU (dual core ATOM). The Xeon PHI is also very competitive in the high performance, low power space and had a significant showing on the latest top500 supercomputer list.The server market is large, complex, and diversified. It requires substantially different features(RAS, high I/O, large memory footprint, ...) than those found in the typical consumer products that ARM fits with today. There is a substantial amount of work that would need to be done to allow ARM to ship into the server market.It is clear we are in the midst of a very large transition (or inflection point if you prefer). It is not clear if INTC will emerge as the X86 monopoly many of us think of it as, or as a world class foundry.I also made a few additional comments about this over on the INTC board here:http://boards.fool.com/i-think-there-are-really-four-possibl...specifically about how Intel fits with Apple.--Alan
Web server market share (Feb 2011):http://news.netcraft.com/archives/2011/02/15/february-2011-w...Denny Schlesinger
It's not often that TMF produces a real article any more, most are just ads for their newsletters and services. Today I had the pleasure of reading a real article which happens to be highly relevant to Intel.Some people believe that Intel will eat ARM's lunch in mobile while others believe the opposite (I'm in the second group). Intel makes what is in essence a standard product in their foundry. ARM allows mobile makers to custom design their chips based on one of several ARM architectures. When using a standard chip you end up with a standard product, not one that differentiates itself significantly from the pack. According to the article "ARM states that there are over 80 licensees for the Cortex family." That's at least 80 firms that are not going to become Intel customers, maybe more.Of course, taking my word for it is a bit of a stretch but if a well known chipmaker decides to exit the field, I think you might take notice. Texas Instruments (TXN) makes the ARM based OMAP 5 mobile processor line and they are exiting the business. They can't compete with "standard" ARM based chips against custom designed ARM based chips. The difficulty for Intel would be even greater unless they had a vastly superior product, which they don't for the mobile market.Texas Instruments and the Big Chip Maker Anachronism By Mark Hibben - November 16, 2012 | Tickers: AAPL, ARMH, INTC, TXN | 0 CommentsOMAP 5 Fails in the Mobile Market PlaceOn Nov. 14, Texas Instruments (NASDAQ:TXN) announced that it was pulling out of the mobile device (phones and tablets) systems-on-chip (SoC) market and cutting 1700 jobs. TI’s principal mobile device products, the OMAP line of SoCs will be redirected to embedded applications such as automotive and robotics. TI management believe they can compete effectively in these areas due to longer product life cycles and lower required investment. Texas Instruments also cited the fact that large customers “are increasingly developing their own custom chips.” http://beta.fool.com/markhibben/2012/11/16/texas-instruments...Denny Schlesinger
Some people believe that Intel will eat ARM's lunch in mobile while others believe the opposite (I'm in the second group). Intel makes what is in essence a standard product in their foundry. ARM allows mobile makers to custom design their chips based on one of several ARM architectures. When using a standard chip you end up with a standard product, not one that differentiates itself significantly from the pack. According to the article "ARM states that there are over 80 licensees for the Cortex family." That's at least 80 firms that are not going to become Intel customers, maybe more. - captainccsYour takeaway from the Fool blog post is flawed in its understanding regarding the competitive distinctions between Intel and ARMH.In brief, TXN is exiting the mobile market because it's determined that its chip, based on an ARM-licensed core architecture, struggles to compete against other chip-makers also using ARM-licensed core architectures with custom overlays. OK. I get that. It's tough to compete in a market where a whole lotta players develop custom chips around a generic core.Here's a good overview of the ARM architecture that's being licensed to whomever wants to pay to play:http://en.wikipedia.org/wiki/ARM_architectureThe ARM architecture describes a family of computer processors designed in accordance with a RISC CPU design developed by British company ARM Holdings. ARM architecture has been in development since 1990 and is the most widely used 32-bit instruction set architecture, in numbers produced.Using the RISC approach, the core ARM processor requires only 35,000 transistors, compared to the millions in many conventional processor chips, resulting in lower power usage and making it very attractive in smaller devices. The company ARM Holdings does not manufacture its own electronic chips, but assigns different licenses to semiconductor manufacturers. It is thus easy for companies to build a low-energy system on a chip for an embedded system incorporating memory, interfaces, radios, etc. The earliest example was the Apple Newton tablet but this same approach is still used in the Apple A4 and A5 chips in the iPad.In 2011, ARM's customers reported 7.9 billion ARM processors shipped, representing 95% of smartphones, 90% of HDDs, 40% of digital televisions & set-top boxes, 15% of microcontrollers and 20% of mobile computers.In short, what we have here is a chip architecture requiring low power mostly as a consequence of the fact that it offers limited computational capability. There's a big market for chips like that, and I expect that ARMH will enjoy healthy revenues for years to come.What Intel offers is something radically different:http://tinyurl.com/cz5p4veIntel® Architecture Leads the Microarchitecture Innovation FieldMicroarchitecture is a blueprint of chip elements. This blueprint, when combined with advanced nanotechnology, enables computing devices to be more capable and energy efficient. Intel’s microarchitecture team continues to make giant leaps in innovation and has recently introduced the world’s first 3D transistors manufactured at 22nm.Introduced in 3rd generation Intel® Core™ processors, Intel’s 3D, 22nm microarchitecture marks a turning point in the fundamental structure of the computer chip. Until now, transistors were 2D (planar) devices. Intel’s 3D tri-gate transistor uses three gates wrapped around the silicon channel in a 3D structure, enabling an unprecedented combination of powerful performance and ultra-low power consumption.Ergo, the difference. ARMH licenses a 2-decade old chip architecture that is cheap to manufacture and can be customized by anyone interested in doing so to suit particular low energy consumption/low computational power needs.Intel has developed a 22-nm 3-D chip architecture that offers low power/high computational capabilities. There's no point in Intel offering licenses to other chip foundries because other foundries simply can't produce the same sophisticated chip architecture. (Incidentally, Intel is already moving towards 14-nm architecture). As I've mentioned several times before, Intel is a full generation ahead of any other chip manufacturer.The only question, then, is who has a competitive advantage...and in what applications? ARMH benefits greatly from offering a cheap chip design well-suited for a myriad of applications. I applaud the company for that. I wish it well.On the other hand, I happen to be one who believes that there is a relentless drive to make enterprise servers more efficient, personal computing devices ever more capable and "smartphones" ever smarter. Companies wishing to offer ever more sophisticated and powerful computational devices may very well opt to turn to Intel for its chip-making expertise, working with the company to produce chips that serve powerful needs. Microsoft did exactly that during the "Wintel" era. In the enterprise space, VMWare (data virtualization) is working with Intel to optimize enterprise server chips.Smartphone makers had been satisfied with ARM-cores for many years. Will they continue to be satified to base their designs around a chip offering limited capabilities? Some, undoubtedly, will. Others? I'm kinda thinking there's a growing interest in developing really, REALLY "smart phones" as well as potent tablets, ultrabooks and all manner of devices I haven't yet imagined.I view ARMH and Intel as two distinct companies offering two very distinct products. I believe the marketplace has ample room for both. It ain't a zero-sum game.
While I agree that ARMH and INTC are very differently positioned companies, I think some of your comparison is flawed in some particulars. The difference in approach between CISC and RISC is not more powerful vs less powerful, but a difference in approach. While the war was "won" in favor of CISC for the desktop, RISC machines such as Solaris and Power still get used for quite high end servers. Moreover, as time goes on, processors get more and more powerful. There are a few applications in which "as powerful as possible" is important, but a whole lot more where "powerful enough" is sufficient. While the demand for power keeps increasing, much of the sales differentiation of the product comes from issues other than cpu power including software features, battery life (aka power consumption), etc.
The difference in approach between CISC and RISC is not more powerful vs less powerful, but a difference in approach. - tamhasOK. Whatev.Here's something to pondercate:http://tinyurl.com/cr4g6dvThe majority of today's processors can’t rightfully be called completely RISC or completely CISC. The two textbook architectures have evolved towards each other to such an extent that there’s no longer a clear distinction between their respective approaches to increasing performance and efficiency. To be specific, chips that implement the x86 CISC ISA have come to look a lot like chips that implement various RISC ISA’s; the instruction set architecture is the same, but under the hood it’s a whole different ball game. But this hasn't been a one-way trend. Rather, the same goes for today’s so-called RISC CPUs. They’ve added more instructions and more complexity to the point where they’re every bit as complex as their CISC counterparts. Thus the "RISC vs. CISC" debate really exists only in the minds of marketing departments and platform advocates whose purpose in creating and perpetuating this fictitious conflict is to promote their pet product by means of namecalling and sloganeering.
Your takeaway from the Fool blog post is flawed in its understanding regarding the competitive distinctions between Intel and ARMH.putnid I could say the same about your post but that would only get us into a religious war which neither one of us needs. LOLARM and x86 architectures are both decades old but neither company has stood still. Their development has gone into different directions to satisfy their respective markets. Besides, computer processing is fungible: just about anything you can do in hardware you can do in software; anything you can do with an ultra fast chip you can match with parallel or multi-core processing. Suppose someone using an ARM license needs a very fast chip, they could put ten cores on the chip and share the onboard memory space.A more important factor than raw speed is power consumption in untethered devices. The chip's physical construction affects it but so does the system architecture. Most computational devices don't operate at full speed all the time. One way to reduce power consumption is by turning off parts of the chip that are not in use. Another is to have multiple cores on the chip and to turn off the ones not in use. You could have a 32 bit core for use when the device is in sleep mode and a 64 bit one for heavy duty lifting. While Intel has concentrated on processing speed ARM has concentrated on low power consumption while increasing the processing capacity. Lately they announced a 64 bit architecture suitable for web servers. As Google has shown, you don't need the fastest servers if you have lots of them. Slower parallel computing can do what faster serial computing can to.At the end of the day all the technical stuff can't tell you which architecture will triumph here or there. You really need to run benchmarks and other tests to find out which architecture is best for your application. In addition there is the business model, not everyone is happy to depend on Intel as sole source of chips. Fool Dirty Dingus called ARM the "kinder, gentler gorilla" for good reason. I believe that Apple is the smartest kid in the room. They traded the IBM/Motorola Power PC for Intel chips in Macs but they went with a proprietary ARM based design for their mobile stuff. Samsung, who makes the Android phones, also uses ARM architecture chips. I can't see any good reason for them to switch to Intel. On the other hand, I have read about several initiatives to build ARM based web servers which would be a direct attack on Intel. I certainly don't say Intel is doomed. All I say is that PCs and servers are Intel's market to lose and they might.Denny Schlesinger
While there has been some convergence ... or perhaps more accurately, cross-fertilization ... this is not to say that they are the same ... and, indeed, you were telling us that they were different in contrasting Intel and ARM.
Hey Denny...You had me with you right until this line...While Intel has concentrated on processing speed ARM has concentrated on low power consumption while increasing the processing capacity. From what I have seen, Intel has been almost single focused on power consumption for the past ten years or so.You really need to run benchmarks and other tests to find out which architecture is best for your application.Totally agree...Here are the current Intel phone benchmarks... now you need to show me some ARM server benchmarks:-) All of these phones are based on the Intel medfield SOC (system on a chip).Orange San Diego phone (Europe):http://www.engadget.com/2012/05/31/orange-san-diego-benchmar...http://www.tomshardware.com/news/Intel-Medfield-Performance-...Xolo X900 (india)http://www.anandtech.com/show/5770/lava-xolo-x900-review-the...http://techcrunch.com/2012/04/19/meet-the-xolo-x900-intel-fi...Motorola RAZR i (Europe and South America)http://www.phonearena.com/reviews/Motorola-RAZR-i-Review_id3...Megafon Mint (Russia):http://www.engadget.com/2012/08/22/intel-brings-medfield-to-...(no benchmark data, but should be similar to others)Lenovo K800 (China):http://www.engadget.com/2012/05/30/lenovo-lephone-k800-medfi...ZTE Grand x in(Europe):http://www.engadget.com/2012/08/30/zte-grand-x-in-benchmarks...note the comparison to previous ARM based ZTE phone which was panned.Intels entry strategy here is very interesting. They have done the complete phone design and lined up manufacturers as well. Then Orange (which is like the US verizon or AT&T) can buy the phones from the manufacturer and put the Orange brand on them. This completely cuts out that middle layer of samsung, lg, or apple and the profit flows directly to the carrier. I suspect the sales force will be motivated to move these phones due to more of the profit going to the carrier.You will notice the lack of a US phone, which is due to the lack of LTE in the current phone design. I think this is the forrest that was missed while talking about the tree (x86 versus arm). Intel needs to develop the complete eco system for the phone, and is a bit behind in some places (the radio for LTE), and ahead in others (silicon hive camera hardware)...Here is a detailed look at clovertrail (a beefed up medfield) tablet.http://hothardware.com/Reviews/Intel-Clovertrail-Atom-Z2760-...As I noted earlier, both medfield and clovertrail are built using the Intel 32nm technology introduced 3 years ago for the PC market. What we see in the reviews are products that are competitive, but not world beating. Early 2013 we are supposed to see an LTE based 32nm product, and then later in 2013 Intel will finally move to 22nm with the phone and tablet chips. When we see benchmarks on the 22nm products we will have a much better idea of how competitive Intel will be in this space. Will they only compete on cost, or will they have superior performance and power?Alan
It is clear we are in the midst of a very large transition (or inflection point if you prefer). It is not clear if INTC will emerge as the X86 monopoly many of us think of it as, or as a world class foundry.I also made a few additional comments about this over on the INTC board here:http://boards.fool.com/i-think-there-are-really-four-possibl......specifically about how Intel fits with Apple.--Alan (or others)Excuse my ignorence of the tech of this area. This is, to me, a leading company which is fighting to keep on the "bleeding edge" of the industry, but has a tremendous built-in base that provides cash (on the positive side) but also a "headwind" to taking the company in new directions. Buggywhip company, Microsoft or [h*ll] Apple (which re-invented itself). It has all the creds IMO to be a value stock.So for my simple mind, how WOULD you rate this company moving forward on a 1-10 scale with 10 high for:Staying a leader (if not THE leader) in the industry?.....who can replace them .. certainly not AMDIts 10 year financial safety?How do you rate it's value right now?...... entry price?Other investment considerations?Thanks Alan and all for contributing to this excellent thread!BobRYR Home Fool
Excuse my ignorence of the tech of this area. This is, to me, a leading company which is fighting to keep on the "bleeding edge" of the industry, but has a tremendous built-in base that provides cash (on the positive side) but also a "headwind" to taking the company in new directions.TMFHockeypop Without a doubt, Intel is all you say but that is not the issue. The industry they serve is moving away from under them and "that" is the issue. It's not "who can replace them" as much as "who is serving the bleeding edge of the market" (mobile vs. desktop kind of thing). No one is going to replace Intel on the desktop but ARM has taken mobile by storm and Intel is playing catch up there.There are two books that discuss these issues, Clayton Christensen's The Innovators Dilemma and Living on the Fault Line by Geoffrey Moore of Gorilla Game fame. In synopsis, when Intel was making 16/32 bit chips for PCs, ARM was making 8 bit chips for use in embedded applications such as disk drives, airbags and ABS brakes. For each x86 chip Intel sold ARM sold hundreds of 8 bit chips. Because many of the application were battery operated, power consumption was an important issue for ARM but back then all desktops were plugged into the wall socket so it was moot for Intel to conserve electricity -- the race was on for the fastest chip on the block competing with Motorola and IBM. The extra heat could be removed with a fan using even more power! While Intel was competing with Motorola and IBM making generic micro processors, ARM was competing with in-house designers using using ASIC chips and the like. ARM was absolutely no threat to Intel (as AMD tried to be) -- until Intel decided it wanted a share of the mobile market. By then the 97 pound weakling had put on some muscle! LOLI have followed ARM for almost 15 years and I'm proud to say most of my predictions for the company have come true while disagreeing with industry experts such as Nick Tredennick and even ARM's management. Intel's future is probably best described by Geoffrey Moore in The Gorilla Game. Intel will be a cash cow as it slowly sinks into oblivion, it might take 10 or 20 years but that is its destiny. Moore was not talking specifically about Intel, that is how Gorillas die.Denny SchlesingerPS: While I have been right about ARM I have often been wrong about Intel on account of my bias toward the Mac. My favorite chip was the Motorola 680x0 which lost out to the Intel x86.PS2: While developing a Desk Accessory for 680x0 Mac I found a way to double the memory it could use by changing the base address of the Mac address table.The Innovator's Dilemma: The Revolutionary Book That Will Change the Way You Do Business by Clayton M. Christensenhttp://www.amazon.com/Innovators-Dilemma-Revolutionary-Chang...Living on the Fault Line : Managing for Shareholder Value in the Age of the Internet by Geoffrey A. Moore http://www.amazon.com/Living-Fault-Line-Managing-Shareholder...Nick Tredennickhttp://en.wikipedia.org/wiki/Nick_TredennickCharles Atlas 97 pound weaklinghttp://www.rrmerritt.com/mabelvale/MHS_Images/I_CharlesAtlas...http://www.boingboing.net/images/_images_Charles-Atlas-comic...
Hey Denny...You had me with you right until this line...While Intel has concentrated on processing speed ARM has concentrated on low power consumption while increasing the processing capacity. From what I have seen, Intel has been almost single focused on power consumption for the past ten years or so. Alan Guilty as charged! That's what happens when you ignore the fact that really knowledgeable Fools are reading your posts! I was going to include it but figured no one would catch me! LOLHere are the current Intel phone benchmarks... now you need to show me some ARM server benchmarks:-) All of these phones are based on the Intel medfield SOC (system on a chip). Sorry, I can't, I don't follow the industry that closely anymore.Intels entry strategy here is very interesting. They have done the complete phone design and lined up manufacturers as well. That's what Intel did in the PC business when Wintel had an 80% market share. Some, like Compaq, fought back and lost to those like Dell who went along. But Intel does not have an 80% market share of mobile or smartphones. Of course, there is a market for Intel's product, the carriers whose interest is not providing the best phone experience but grabbing market share with cheap phones possibly as a loss leader. This is again a replay of a strategy that kept Motorola from making heaps of money on the original cell phone. Geoffrey Moore comments on it in The Gorilla Game: "Compaq has been a king in PC servers, having traditionally held a two-times lead over Hewlett-Packard and IBM, although that lead has never been safe. Seagate is a king in hard disk drives, 3COM in Ethernet cards, US Robotics in modems, and Motorola in cell phones and pagers. All of these companies lead their respective markets, but none has the power to manipulate a de facto standard, and none enjoys high switching costs. Thus Motorola, despite its 50% market share in cell phones, is under severe margin pressure from the cellular access providers who threaten to substitute another vendor's product for their free-phone offer to their customers." Page 63.Note that most of these kings are gone. In the Wintel duopoly it is my contention that most of the gorilla power came from "Win" and only in the hardware end did the "tel" have effect. Since then, Apple switched Mac hardware three times (M680x0 -> PowerPC -> Intel) showing there is no hardware gorilla power anymore. In mobile it's Intel vs. in-house design. Carriers will go with Intel. Smartphone makers will go with in-house design. What will smartphone and tablet buyers buy? Figure that one out and you'll make lots of money. Of one thing you can rest assured, the gorilla power of the PC days has been strongly eroded. "Computers" have been commoditized.Denny Schlesinger
Staying a leader (if not THE leader) in the industry? .....who can replace them .. certainly not AMDWe need to start out by defining "the industry" that Intel is in, as there is some disagreement on that. If you look at their website and you look at their history, and you look at what they say it is clear to me that the industry they are in is "semiconductor manufacturing". So, we would need to compare them to other "semiconductor manufacturers". This rules out companies like ARMH, AMD, QCOM, nvidia, and AAPL. If we narrow semiconductors down to "logic" as opposed to DRAM and Flash we see the primary competitors as Global Foundries (the old AMD factories), TSMC, and Samsung. IBM and TI have historically been on that list, but have been "run over" by Intel or others. The perception of some (including Denny) is that X86 is the moat. I believe the moat has switched and become the underlying manufacturing technology. Each new generation of manufacturing technology provides three benefits: reduced power, increased performance, and lower cost. All of those are as important, if not more important, to the phone and tablet market as they are to the PC market. We see this effect in the iPhone product. The original iPhone has a cost to Apple of around $250, while the new iPhone 5 is closer to $200... with much higher performance. Speaking of the iPhone, looking at the players and cost sheds some light on the value chain.ARM provides the cpu design for both the baseband chip and the application processor. Annual ARM revenue is $0.76B. Imagination technologies ($.128B annual revenue) provides the GPU design for the applications processor. Qualcomm integrates an ARM core with Qualcomm IP and has TSMC build the baseband chip. Qualcom annual revenue is $19B, and TSMC is around $14B. Apple integrates an ARM CPU with the GPU, and additional Apple IP and has samsung manufacture the Ax chip for the phone. For the $600 iPhone apple gets $400, qualcomm gets about $35 which TSMC gets part and ARM gets part. Samsung gets around $17. The various other suppliers and assemblers get the remaining $150 or so... very rough numbers but it is clear this is not a really stable situation and it is unclear how it will evolve.Its 10 year financial safety?The first question I ask here is what is happening to the overall semiconductor market that Intel serves? Today they are about $50B, with most of that revenue coming from the PC at $100/unit times 400Mu/year.If we look at the phone market the applications processor is about $15 but ships close to 2Bu/year for potential revenue of $30B, still shy of what INTC makes on the PC market even if they garner 100% of the market. This makes things look pretty risky for Intel unless you assume the existing PC market at least stays stable.If we ask a different question though we get a different answer. What is happening with the worldwide demand for compute capacity? I think the answer is that it is going up, but the form factors it is delivered in is changing. Intel has a semiconductor capital manufacturing base of about $27B, with an annual capital expenditure of around $11B. TSMC has an installed based of about $14B and capital spending at $7B/year. It seems Intel is one of the few positioned with factory capacity to meet the worlds demand for compute capacity, and that is not going to change very fast.A worst case scenario is to assume all Intel IP is of no value. TSMC makes $14B/year on a capital base of $14B (approximately). This means INTC revenue would bottom at their capital base or $30B versus the current $50B. I really don't see that happening, but would consider it the worst possible outcome.How do you rate it's value right now?...... entry price?Given their silicon manufacturing technology lead, Intel can sell silicon at TSMC cost, and still make about 30% margin (versus current margin of above 60%). The stock is significantly discounted (4.4% dividend yield, PE of 9) due to a reduction of revenue in the current transition. It is clear to me that the current down trend is not permanent, nor even sustainable. The questions remain about how low it will go, and when will it turn.Other investment considerations?In the value chain for mobile devices I really like Qualcomm.As a final note on the underlying silicon technologies, we hear lots of buzz words like 3d transistors and "high K, metal gate". These technologies typically take over 10 years from first conception until you buy them in some product or other. Staying on the leading edge here is so expensive companies started combining efforts, and still failed. I really don't see how Intel can NOT be the dominant silicon in future computing devices. With a ten year or more lead time on developing the underlying technologies it substantially rules out new players.Alan
Hi Denny,A few thoughts on your post...now you need to show me some ARM server benchmarks:-) ...Sorry, I can't, I don't follow the industry that closely anymore.The reason for the smiley is that there really are not any ARM servers.That's what Intel did in the PC business when Wintel had an 80% market share. I see this as a bit different.Apple builds the iPhone at Foxconn, for about $200. They sell this phone to AT&T for $500 or so, pocketing the profit. AT&T turns around and sells it to their customers for $200, and makes up the $300 on the service contract. Intel is giving AT&T the option to buy a phone directly from Foxconn for $200.Of course, there is a market for Intel's product, the carriers whose interest is not providing the best phone experience but grabbing market share with cheap phones possibly as a loss leader.You are making some assumptions here about phone quality that I am not sure are valid.Carriers will go with Intel. Smartphone makers will go with in-house design. I am not really sure what this even means.Alan
Good stuff, Alan. I appreciate your detailed insights. Thanks.
a second on that thanks to Allen--incredible knowledge
The reason for the smiley is that there really are not any ARM servers. You are not looking hard enough. ;)I see this as a bit different. ... Intel is giving AT&T the option to buy a phone directly from Foxconn for $200. We agree on that. But it's an inferior product...You are making some assumptions here about phone quality that I am not sure are valid. ...hardware-wise it's a perfectly good product. But it's no iPhone. Maybe not even an Android. What differentiates these phones is not the hardware but the software, the OS, the human interface. When Intel sold PC motherboards it counted on Microsoft to provide the software. The Orange Intel phone is Android:Orange San Diego review: Intel does phones, finallyhttp://www.engadget.com/2012/06/14/orange-san-diego-review/Despite all the technological wonders of Intel's Fabulous Fabs, which I don't ever deny, it's still the software that really counts.Denny Schlesinger
it's still the software that really countsWhich comes back to the point that for many users and applications, hardware that is "good enough" is just that and other features will make the purchase decision.
Maybe not even an Android. What differentiates these phones is not the hardware but the software, the OS, the human interface. When Intel sold PC motherboards it counted on Microsoft to provide the software. The Orange Intel phone is Android:...A little bit rambly, and perhaps trying to talk yourself into or out of something. All the Intel phones are android today. It is not clear if microsoft will do a windows 8 phone on Intel or not. The current windows 8 phones are getting really good reviews, and run on ARM.Given Android currently out ships iOS 2 to 1 it seems having Android on Intel phones right now isn't a bad thing. I didn't really mean to start an Apple holy war, but the iPhone 5 is the most readily available build costs. I suspect the Samsung galaxy S3 would be a similar cost ($200) and sells for about $700 so about the same model as with the iPhone.Despite all the technological wonders of Intel's Fabulous Fabs, which I don't ever deny, it's still the software that really counts.I would argue everything counts. The hardware makes a difference on weight, battery life, and responsiveness. Those are not software things, and they are things that matter to customers.The cpu and GPU performance in phones has grown steadily over the past five years. Are we suddenly declaring it doesn't need to grow any more?I'll end with a bunch of questions that will hopefully be answered over the next 6 to 12 months.(1) What is going to happen with tablets? The surface RT (using ARM) is getting panned. Windows 8 looks very strong. Intel is late with their clovertrail chip for tablets. How will the Intel Haswell chip which should ship next March or April do in the tablet market? will iPad and iPad mini remain dominant in this market?(2) will microsoft port windows phone 8 to X86?(3) how will the LTE phone chip Intel ships in early 2013 perform relative to the competition.(4) When does Intel finally move phones to a more leading edge 22nm technology? I expect October 2013.(5) what part will the new ultrabook / tablet convertibles play. I personally really like both the Lenovo Yoga, and the dell XPS 12.http://www.dell.com/us/p/xps-12-l221x/pd--Alan
1. First, great series of posts. Thank you.2. To these other questions ... it seems to my incomplete understanding that similar questions have been asked for maybe three decades now. So far Intel's leadership and their ability to keep great engineers has won. But, as you suggest, it is a cyclical industry and their defined competition keeps changing.3. Again, excellent.Bob
I'll end with a bunch of questions that will hopefully be answered over the next 6 to 12 months. The big problem for the investor is that high tech markets are really difficult to predict, unlike commodity markets, cost or even quality are not the ultimate arbiters. The winner is picked by the market as if by chance. Once a winner is picked, displacing it is very difficult. The usual way to displace a previous winner is by way of "disruptive technology" which essentially is a much less capable technology at a fraction of the cost which creates a new market. As this technology goes upscale it starts to eat into its predecessor's markets. ARM-Intel happens to be a perfect example.The market will answer your five questions in its own good time. BTW, I don't have a position in any of the companies we have discussed. I've put high tech on the "too hard pile."Denny Schlesinger
Intel President announces retirement. Price declines substantially. Value play at 9:30 am?http://newsroom.intel.com/community/intel_newsroom/blog/2012...Hockeypop
Maybe it's Otellini...Intel CEO Paul Otellini to retire in May Intel CEO Paul Otellini plans to retire in May; board will look for his replacementAssociated Press – 15 minutes agoSANTA CLARA, Calif. (AP) -- Intel says its CEO, Paul Otellini, plans to retire in May after nearly 40 years with the company.Intel Corp. said Monday that its board will look at company executives as well as external candidates to replace Otellini. He has been head of the world's largest chip-maker for eight years. http://finance.yahoo.com/news/intel-ceo-paul-otellini-retire...Denny Schlesinger
It is Ortellini. For some reason I got the wrong press release, althought it's interesting that several top exec's are leaving.http://newsroom.intel.com/community/intel_newsroom/blog/2012...Hockeypop
Alan is spot on w/ regard to the INTC mfg technology.Anyone here think the likes of QCOM wouldn't swamp INTC fabs with orders if INTC stepped back from mobile? Hah!There'd be a line of fabless semiconductor companies out the door wanting to bid for and allocation of that INTC capacity.In the early days (90's) of QCOM's fabless chip adventure, INTC was QCOM Supplier of the year several times. Then INTC decided to get into the mobile chip business and buy up long since forgotten wireless semi companies they thought would let them catch up to QCOM on the cheap.
And interesting new twist on ARMhttp://phys.org/news/2012-11-samsung-biglittle-processor-iss...Two sets of 4 core processors together - one low power, but energy efficient, and the other high power, but used only when needed for performance.
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |