INTC - Intel

[ItsAValueTrap]

Hey does anybody have any insight on TSMC?  Right now they aren't that competitive against Intel.  However, the future could change dramatically.  If they get to a big enough size, their manufacturing technology could be competitive with Intel's.  Their capex of about $7.3B this year versus Intel's $10.7B will kick into effect in the future.  If Intel doesn't take a lot of market share away from the other smartphone SoC companies soon, maybe it is possible that TSMC will grow to a size where their manufacturing is competitive with Intel's.  Then AMD will be able to compete with Intel on more equal footing... this would be very bad for Intel.

 

Currently TSMC's best technology is 28nm without finFET and they are having yield issues.  Intel is shipping 22nm parts with finFET.  They are actually overcapacity at 22nm (I'm guessing partly due to very good yields) and can't migrate their 32nm designs down right away because it takes time to migrate designs.  Intel has a pretty big lead right now.

AMD is in really bad shape financially and was heavily shorted last time I checked.  Their product roadmap does not look competitive versus Intel.

[Guest valueInv]

Some momentum at ARM:

 

http://gigaom.com/cloud/facebook-amd-hp-and-others-team-up-to-plan-the-arm-data-center-takeover/

[ItsAValueTrap]

I don't think ARM is a big threat in the server space at all.  AMD x86 is the biggest threat (AMD actually pioneered 64-bit x86 in servers), followed by all the other RISC guys (SPARC/oracle/sun, Power/IBM, ARM).

 

Here's my blog article on it...

http://glennchan.wordpress.com/2012/10/14/arm-intel-and-amd-in-the-microserver-space/

---------------------------------------------------------------------------------------------------------------------------

 

ARM, Intel, and AMD in the (micro)server space

 

October 14, 2012 by glennchan | Edit

 

There is talk about ARM making inroads into the server market.  I don’t believe that this will be a huge threat to Intel.

 

Wimpy cores versus brawny cores

 

Calxeda and Tilera are two ARM-based products that offer are a “wimpy” core approach.  For the same level of work performed, a large number of less powerful processor cores will use less power than a smaller number of more powerful processor cores.  Since power is a huge cost in running data centers, less power consumed may translate into lower overall costs.

 

However, the wimpy core approach is unsuitable for many applications.  The Google paper “Brawny cores still beat wimpy cores, most of the time” explains why.  The short story is that many problems map poorly to wimpy cores. The corollary of the paper’s title suggests that there are niche applications where the wimpy core approach is the best.

 

Intel has a wimpy core solution with its Atom and Centerton processors.  The ARM-based solutions take the wimpy core approach to an extreme compared to Intel.  Their products are far “wimpier” than Intel’s products and may do well in a niche within the wimpy core niche.  For a very large number of extremely wimpy cores, the ARM-based solutions may have an advantage over x86 cores as each x86 core will have some overhead related to the x86 instruction set.  Within this extremely small niche, I am not sure if Calxeda and Tilera have had much success in gaining traction in the server market.

 

ARM-based server products currently have low rates of adoption.  Facebook may or may not choose ARM in the future: “Facebook is evaluating ARM’s chips for their low-power characteristics and reasonable compute capability. However, it won’t be until the launch of 64-bit variants next year that it gets serious about the processors, the engineers indicated.” (http://www.zdnet.com/facebook-stretches-arm-chips-in-datacentre-tests-7000004707/)  HP’s Project Moonshot was originally designed with ARM-based processors in mind but will first ship with Intel’s Centerton.

 

Seamicro

 

Seamicro is a startup company (purchased by AMD) that produces an innovative solution for the server market.  In a typical server, each CPU needs to be supported by a motherboard with various chips on it (e.g. networking, storage) and at least one hard drive.  Seamicro’s solution allows multiple CPUs to share all these resources.  This produces savings on hardware costs as you need a fraction of the networking and storage chips for each CPU and a fraction of a hard drive.  Power costs also go down since a lot less hardware is being powered up.

 

Seamicro’s technology plays well with the wimpy core approach.  One problem with the wimpy core approach is that less powerful processors means that you need more servers… leading to higher hardware costs.  Seamicro’s approach reduces hardware costs and shifts the economics more in favour of wimpy cores.

 

AMD and Seamicro will likely pose far more competition to Intel than ARM-based servers.  I expect that AMD will design processors that play well with Seamicro’s product.  Currently there is a Seamicro product that uses Intel Xeon CPUs with many features on the Xeon CPU turned off (e.g. SATA, integrated graphics, etc.).  The obvious evolution is for AMD to produce a similar chip without these features.

 

Semiaccurate has a great article on AMD’s purchase of Seamicro.  While the author portrays AMD’s purchase as a huge loss for Intel, his predictions about the future haven’t always been accurate in the past (e.g. Facebook using ARM servers; the “tectonic shift” didn’t really materialize).

 

Intel Centerton

 

Intel’s approach to the microserver market is to basically put an entire server on a chip.  The current trend in all areas of computing has been to do a “system on a chip”.  Instead of using many different pieces of silicon, manufacturers have been merging everything together onto a single piece of silicon.  This saves on packaging costs and saves space in smartphones.  Whereas Seamicro has each CPU sharing networking chips (with a chip facilitating the sharing of resources), Intel simply combines the CPU and networking chip on a single piece of silicon.

 

I couldn’t find much data on the relative costs and performance between Centerton’s approach and Seamicro’s approach.

 

Where the future is headed

 

Intel sees microservers growing but it expects them to take up 10% of the overall server market by 2015 from 1% to 2% today.  From Intel’s latest 10-K, the data center group had revenues of $10,129 so microservers might only be a $1B/yr market for Intel.

 

I think that ARM-based products will not pose a serious competitive threat to Intel.  Even if they are successful, these products are only currently targeting a niche within the microserver niche.

 

In the overall server market, RISC-based products (e.g. Power, Sparc) will likely continue their trend towards lower market share.  Intel’s Itanium CPU, which does not use the x86 instruction set, will lose share to Intel’s x86 Xeons over time as Xeons are just as fast and will offer the same high-reliability features.  AMD will likely be Intel’s most dangerous competitor as it has high volume and an x86 license.

[Guest rimm_never_sleeps]

amd will be lucky not to go bankrupt. tech investing is about looking to the future not the past. which big customers are going to Trust AMD with mission critical services? ARM is where the puck is going. Amd is where it was.

[ItsAValueTrap]

Um... AMD currently has more Reliability, Availability and Serviceability (RAS) features than ARM.

 

Where ARM will go first is not in mission-critical services (which is server by Power, Sparc, Itanium, and AMD/Intel x86).  ARM will target the wimpy core niche first.  Based on vendors' currently-announced plans, that is where it is going.  And of course Intel is attacking wimpy cores, smartphones, tablets, etc.

[Palantir]

Isn't AMD partnering with ARM for the 64 bit servers?

 

Anyone have a good breakdown of the types of servers in the broader server market? I think that will give us a clearer picture of where the companies are fighting for market share.

[Guest rimm_never_sleeps]

again I don't care what the specs say. if you are a CIO and you rely on AMD you should be fired. amd has been the most unreliable tech company going back to the 80s. they have created zero shareholder value going on three decades. in fact I just read that AMD was investigating getting an ARM license.

[ItsAValueTrap]

rimm_never_sleeps:  haha.

 

Anyone have a good breakdown of the types of servers in the broader server market? I think that will give us a clearer picture of where the companies are fighting for market share.

 

-High-performance computing, supercomputers, etc.

There are some markets where it is difficult to increase performance by increasing the # of servers.  So they need really fast servers.  Each server will have multiple sockets (2, 4, etc.).

 

On the lower-end, you might have people with 2-way/2-socket servers to host forums and things like that.

 

-Mainframes are similar to HPC?

 

-Massively parallel problems

You can increase performance simply by adding more servers.  Most cloud applications fall in this category (email, serving web pages, etc.).

 

Google is like this though I'm not even sure if they even use server hardware.

 

-Wimpy cores

Within this market, there is a wimpy core niche.  Some problems can run on servers with slower but more cores.  This is a more power-efficient approach that lowers TCO.

Some people call these microservers and expect the microserver market to be 5-10%.

 

-Servers with GPGPUs

 

 

In terms of features and hardware:

-How important single-threaded performance is versus multi-threaded performance

-Do you need virtualization

-Do you need VNC

-Do you need the highest RAS features

-Does each individual server need to be a certain speed.  Some people need really faster servers and pay a huge premium for the extra performance.

-Do you need 64-bit

 

[FrankArabia]

I find all this tech talk about processors/servers etc rather misleading regarding the future business of Intel. Firstly, its backward looking. Industry leader here today gone tommorrow.  Second, lots of the pluses/minuses are very subjective. I am focusing on industry economics and the cost of reproducing assets instead.

 

As a case in point, it took me almost 2 weeks to find out that intel has/can make chips just as efficient as ARM licensed chips.

[Ross812]

I doubled down on Intel today. It now makes up about 11% of my portfolio.

[FrankArabia]

I doubled down on Intel today. It now makes up about 11% of my portfolio.

 

I have a smaller position (4%) but learning about this biz. I might go in a bit more but I need to more certain on the economics of the business. I personally feel the whole ARM problem is way overblown but still....

[Ross812]

I have a smaller position (4%) but learning about this biz. I might go in a bit more but I need to more certain on the economics of the business. I personally feel the whole ARM problem is way overblown but still....

 

Revenue is down for the last quarter ~20% yoy. Analysts have Intel EPS growth down 6.6% next year. Intel is having a rough year, it's that simple. No one knows how well Windows 8 is doing right now, and IMO people put off purchasing a new PC until W8 came out and all the new PCs where released. I know I did.

 

I would not worry too much about ARM taking any market share from Intel in the server space. Intel cores cannot be matched by RISC architecture. Intel has been growing in the server space by 6.5% and the trend should continue. RISC architecture is best used to host forums, email, and cloud storage. If you see Internet based TV, video, games, shopping, and animation rich advertising growing in the future, it will be on the back of Intel built processors. Lets see how risk processors have faired in the server space for the last 10 years:

 

http://static.cdn-seekingalpha.com/uploads/2012/10/17/1095245-13504499641933699-Ashraf-Eassa.png

 

Now for the PC segment and the 'Death of the PC'. The worst report I could find is the new Barclays report:

 

Reitzes cut his 2012 forecast to 352.75 million PC units sold, down from 363.92 million, cut his 2013 model to 338.34 million, a 4% decline, down form 362.6 million previously, cut his 2014 model to 324.1 million from 360.1 million, and cut his 2015 model to 311.5 million from 351.4 million.

 

Intel's business as a whole on the 2011 Annual Report:

 

http://static.cdn-seekingalpha.com/uploads/2012/10/17/4222561-13504516179627302-Ed-Liston_origin.jpg

 

At 54B (ttm revenue) we get:

 

PC segment - 38.9B

Server segment - 10.8B

Software (McAfee) - 2.2B

Tablet/Phone (atom) - 1.1B

Storage (SSD) - 1.1B

 

So if we project our worst case scenario of a 4% PC decline:

PC Revenue 2013- 37.3B; 2014- 35.8B; 2015- 34.4B; This is about 1.5B a year in lost revenue. Servers (higher margin) are growing at about 0.75B a year.  I feel very conservative saying the SSD segment will grow at 30% a year (since analysts are forecasting saying 63%). This comes to .44B in revenue growth. 

 

So an investment in Intel depends on the revenue growth of 310 million dollars a year in the Tablet/Phone space. (-1.5B PC's + 0.75B Servers + 0.44B SSD = 0.31B Tablet/Phone to make up for PC loss). At todays prices, Intel is priced for no growth going forward. To make a better return than the 4.4% dividend right now, Intel has to grow their atom group at greater than 28% going forward. Keep in mind the first atom tablets just shipped and the current revenue from their atom segment is from the netbook market. The tablet and smartphone market is growing at 60% with the atom being able to run both Android and W8.

 

Intel's growth really does come down to the mobile market. Intel has dominated every market they have entered. An investment in Intel is a bet that Intel can grow their Atom segment at at least half the rate of their competitors. Any other gain: Intel becoming very successful in the mobile market, increased server segment growth, or a better than current worst case PC forecast I referenced (most analysts have the PC market growing at 2%-3%) is just icing on the cake. 

[Guest rimm_never_sleeps]

the reason intel has beaten risc in the past is not because of a better architecture. it's because the most ubiquitous software ran on intel. intel had the ecosystem. in fact most analysts noted that risc had a much more efficient architecture than x86. intel won because they had the numbers and because they had msft. the situation is different now. arm has the numbers in mobile. arm also has msft goog and appl now. there are reports that apple is preparing to move away from intel. goog just released an arm based chromebook that is almost as good as the intel one, yet far cheaper.

 

yes intel has dominated in PCs, mostly thanks to inept competion and msft. but that's in the rear view mirror. you fail to mention intel's abysmal record of entering markets via acquisition at cycle tops. the Mcafee acquisition is just baffling. the worst case PC market is way worse than down 4%.

 

having said all this I would not sell intel here. I would not buy intel here. if I owned I would hold it. It's closer to a buy than a sell for me.

[mikazo]

arm also has msft goog and appl now. there are reports that apple is preparing to move away from intel. goog just released an arm based chromebook that is almost as good as the intel one, yet far cheaper.

 

I'm not too worried about Windows-on-ARM being the Windows-on-x86 killer, for the following reason:

(Quote from Wikipedia)

Only Windows Store apps can be installed on Windows RT—most Windows Store apps are cross-compatible with both x86 and ARM architectures. Developers will not be able to create or port applications to run on Windows RT using the Win32 APIs—meaning that existing "desktop" applications compatible with x86 versions of Windows cannot be used on Windows RT. Microsoft developers blocked the use of third-party desktop applications on Windows RT because their inclusion would prevent the operating system from meeting its goals of "longer battery life, predictable performance, and especially a reliable experience over time."

 

I sort of doubt that the Chromebook will take off either. I haven't met a single person that's bought or used one.

[Guest rimm_never_sleeps]

yeah I "sort of" feel the same way about Intel trying to get into smartphones and tablets. don't know a single person who has a phone or tablet with an intel chip in it.

[Ross812]

The worst case PC market is way worse than down 4%.

 

Please reference a lower prediction for Global PC sales than the Barclay's -4%.

 

There are reports that apple is preparing to move away from intel.

 

Please cite a single report that is not from anonymous sources. The article I read said an anonymous Apple engineer said they were working on an ARM chip that could power a laptop. Seeing how the Atom z2670 (based on a 5 year old platform) benchmarks well with the Apple's new A6 processor; I would not be too worried. The Silvermont Atom with a brand new architecture will be released in the second half of 2013. Please think for a second about the 'nonsense' article about Apple leaving Intel. Apple recently switched to Intel because Intel Performance and Supply was too compelling to stay with PowerPC. Furthermore, the i7 mobile processor in the MacBook Pro is roughly 13 times more powerful than the A6 processor. Next years Haswell processor has been demonstrated to run a 7 watts at full utilization, and 10% faster than Ivy Bridge. Now I know Apple can do some amazing things but they are not going to get anywhere close to 13x the performance out of their ARM processor while increasing the power by only 4.5x. This may be possible someday but never in Q2 of 13. I'm also willing to bet Intel does not sit still and continues to develop their chips past Haswell...

 

[Guest rimm_never_sleeps]

why on earth would one believe that the "worst case" for the PC market is the lowest sell side estimate? the sell side is known for huddling around the estimates of their peers. do you really invest that way or do you sometimes think for yourself? I made no assertion that apple would leave intel. I merely noted that there are reports out there that this is a possibility. You're really emotionally wrapped up in this investment it seems.

[FrankArabia]

I agree, do not use SS estimates at face value but more as a gauge or check. however, 4% y/y is not ridiculous though one should position for a much worst situation.

[ItsAValueTrap]

The whole RISC versus CISC debate is mostly irrelevant today.  CISC CPUs need to dedicate extra transistors to decoding the CISC instruction set.  Because transistor counts are so high now, the burden of those extra transistors is negligible.  Intel's Itanium chip is based on a VLIW architecture which came after RISC (it didn't work as well in practice as hoped since it was too hard to write the compilers).

 

yeah I "sort of" feel the same way about Intel trying to get into smartphones and tablets. don't know a single person who has a phone or tablet with an intel chip in it.

The Intel-based smartphones don't support LTE and none of them are being sold in the US.

 

Generally speaking, tech companies have to make bets on technology a few years ahead of time since that's how long product development takes.  Intel is making a push into smartphones and tablets now... they bought Infineon wireless a few years ago.  Currently their push into smartphones has only been semi-serious as they didn't use their best fabs to make smartphone SoCs.  Smartphone SoCs have a lot less die area than CPUs so they actually aren't that powerful (and dedicate a lot of the area to non-CPU stuff like wireless, image processing for the camera, and hardware video encode/decode).

 

Currently, Intel's chips are middle of the pack and Intel may be at a slight manufacturing disadvantage compared to its competitors.  Its smartphone chip is built on a 32nm process versus 32/28nm for most competitors.

Intel should ship chips based on a 22nm process next year and use its manufacturing advantage for smartphone SoCs.

[Guest rimm_never_sleeps]

you're right. intel should have seen the wireless opportunity years ago. intel totally missed the wireless opportunity. they are attacking it now. but where are we in the wireless cycle? ti already has exited the SOC business because they can't make any money at it.

[Ross812]

why on earth would one believe that the "worst case" for the PC market is the lowest sell side estimate?

 

I used the worst case estimate “I could find”. What is your recommendation for a worst case estimate? Remember the majority of growth in the PC market is coming from developing economies purchasing work stations. The upgrade cycle may be extended because consumers are increasingly using tablets in place of laptops for everyday computing in the home, but businesses (both in developed and developing economies) rely on laptops, ultrabooks, and desktops. Most forecasts for the PC industry are for modest growth. What decline should be assumed? Should you add 50% to the downside and say 6%? Assume an average of a 2.2B/yr revenue decline? What are your reasons for assuming lower than a 4% decline? If you are going to comment add something constructive.

 

the sell side is known for huddling around the estimates of their peers. do you really invest that way or do you sometimes think for yourself?

 

I try to approach investing by looking for good businesses with a competitive and placing them on a watch list. I read the annual reports for the businesses on my watch list, and stay current with any articles written about them. I come up with a cursory buy price by looking at the fundamentals, analyst estimates for growth earning ect., DCF, and so on. It depends on the type business as to what metric I use to determine its value. When the stock approaches my cursory buy price I dig into the fundamentals and try to handicap growth rate, the balance sheet, and any obstacles the company faces. What I laid out in the previous post was a back of the napkin look at what would happen if the PC market declined at 4% a year. TBH, thinking for myself, I think the PC market is going to stay at +/- 2% growth for the next few years. Ultimately, I see phones having docking stations with a mouse, keyboard, and monitor (and even laptop like docks).  I don’t see this as an earth shattering move, but gradual as phones get more and more powerful. I’m also willing to bet that Intel is so far out in front that they can reduce their power from the Hasswell chip by 70% (as Ivy to Hasswell was a 60% reduction) before any of the ARM manufacturers can increase the processing power of their chips by 1300% while maintaining a TPD of 2-3W.     

 

I made no assertion that apple would leave intel. I merely noted that there are reports out there that this is a possibility. You're really emotionally wrapped up in this investment it seems.

 

If you don’t believe Apple would leave Intel then why talk about it? If you don’t want to be called out for citing BS reports then don’t call attention to them. It seems apparent from your posts that you read a lot of financial news, but you don’t seem to back check your info very often. I’m not emotional about the investment. I’m calling BS on the Apple leaving Intel report and gave a reason. I also asked for a worse sell side report than the one I used. If you have a reason to think -4% is unrealistic or have a link to a credible report, I’m all ears. I would like to have some good discussion about this , as it really applies to MSFT, HP, and Dell too. Referencing BS articles, saying I think 4% is way too low because sell side analysts are over optimistic and follow the herd is not constructive. 

[Guest rimm_never_sleeps]

you may not wish to hear that "bs" report about apple, because it doesn't fit into your preconceived notions about your beloved intel's perceived dominance. However, investing is not about protecting turf or stifling or discouraging information flow. more information = good. less information = bad. Other less prickly investors may not mind discovering that articles have been written (more than one) about Apple possible making a major cpu switch, a switch that would have a huge impact on intel. In fact that would be something worth investigating no? you investigated and determined (hopefully correctly) that's it's not going to happen. other's may have a different conclusion. this is a forum to Exchange information. 

 

btw, what do you care that a "bs" article was cited? why do people get so emotional about their stock picks? If you seek to squelch dissenting views, you're playing in the wrong sandbox. That's not what great investors do. They welcome and seek out dissenting views, whether it's BS or not. Your investment will work or not work based on facts. Not whether "bs" articles are posted about it.

 

As far as coming up with a worst case projection, it's not "down 4%" for me. That's probably what we have this year, no? my worst case would result from a double whammy of a serious recession in the global economy and a continuing migration away from PCs to smaller mobile devices. Throw in a "vista" like reception of Windows 8 for good measure. A ballpark figure would be down 5%-10%. When building a worst case, you think about everything that possibly go wrong.

 

However, I don't invest that way. I don't care so much about next year. Intel is already discounting a bad year next year. I care about positioning for the long term and free cash flows over the long term. And on that score I don't believe Intel is well positioned at all. Be that as it may, at less than 10 times earnings, investors are probably too pessimistic here.

[ItsAValueTrap]

If you look at history, the growth of the PC market doesn't always mean profits for hardware manufacturers.  Many hardware companies have not made money despite the huge secular bull market in computers.  The thing about the manufacturing industry is that there is often intense price competition that drives profits down.

 

Intel is only finally starting to make more money than the Internet bubble days.

 

Going forward, I think that you would need to guess/predict the price competition between Intel and AMD.  Considering that AMD is in dire financial straits and diversifying into new markets (e.g. server chips with ARM cores), I don't think that they will be aggressive with pricing.  On the manufacturing side, Intel is currently ahead of AMD and its fabrication partners.  If contract fabs like TSMC get really aggressive with capex then TSMC's future customers (AMD, the smartphone SoCs companies that Intel competes against) will benefit.

 

TSMC's biggest competitor is probably Intel.  The majority of TSMC's customers compete against Intel directly.  Intel's biggest threat IMO is TSMC.

 

Intel has dominated every market they have entered.

Actually... they have failed in:

Memory (Intel was originally a memory company)

RISC CPUs

Discrete video cards

FPGAs

Networking chips

Larrabee / the second attempt at a discrete video chip- did not ship

Smartphones before Medfield

 

? Itanium/Itanic - technically it is profitable now, though it will ultimately be cannibalized by Intel's Xeons.

? Flash memory - I don't know if Intel has made or lost money in flash memory.  (It had a joint venture with Micron which it sold to Micron.)

? Embedded

 

Intel has dominated in:

Desktop CPUs

Server CPUs

Integrated graphics, despite many years of atrocious hardware and software drivers

[ItsAValueTrap]

The idea that Apple would switch their CPUs to ARM is pretty crazy.

 

There would be a burden on Apple to support two different sets of hardware and associated software again, like what happened when they switched from powerPC to Intel.

 

Switching to AMD would make more sense- apparently AMD was very close.  But the problem with switching to AMD is that somebody would need to write the drivers for all the hardware associated with AMD CPUs and video cards.  And Apple actively tries to avoid too many hardware configurations out there as it makes software more difficult to develop and the computers more buggy.  Their strategy may not be the correct one but it has generally been their strategy in the past.

 

Plus, Intel CPUs are better on the high-end.  So I really doubt Apple would switch.

[Ross812]

rimm_never_sleeps,

 

I don't have a problem with free flow information in the least. I have a problem with articles from anonymous sources that take half a brain to realize they are untrue. Please go to the BAC thread and cite some anonymous articles about how Bank of America is destined for bankruptcy, continue to defend your article, and see what happens. I'm not squelching a bear case view; I am telling you why it is implausible and calling to attention that there is no source.

 

Please look at Itsavaluetrap's posts above to see some decent discussion. I've looked back at 5 pages of your posts and can't find anything that suggested you read anything more than the news feed on yahoo finance.

 

As far as coming up with a worst case projection, it's not "down 4%" for me. That's probably what we have this year, no?

   

 

No. -1.2% (http://money.cnn.com/2012/10/10/technology/pc-shipments-decline/index.html)

 

Thank you for your explanation of why you are estimating -5 to 10%. I'll stick with the Barclay's analyst. Hopefully he did more homework. If you care to actually look into Intel or the sector in general. Feel free to comment.