Intel (INTC) critics often say that the company can only succeed in a competitive void. Apparently they haven't studied the server chip battle.
Intel actually started late and well behind in the server business. They then made some humongous mistakes … umm such as Itanium. In the end the Intel Xeon family crushed all comers to dominate the server market with 95% of server units and a mind boggling 97% of server dollar share.
"Here's a bit of an eye-opener for you. The high-end server market is tiny compared to the PC consumer space, and the RISC market SPARC plays in is a tiny fraction of that. 9.5 million x86 servers shipped in 2012. The number of RISC servers (including IBM's Power, Itanium, and SPARC): 154,870."
The server competitors were no toy companies; Oracle, IBM, and AMD who had rights to the X86 microcode and was a player in the PC CPU business.
Intel, on the offensive, focused on a goal is not a company to take lightly. I might take a moment to point out to the RISC/CISC debaters that two of the victims of the Intel server offensive were RISC chips, actually three if you want to count the Itanium.
Yes, Intel even ate their own baby, Itanium, to win the server battle. In the end, Intel convincingly won the server part of the mobile business.
To be fair, I can also remember Intel's failure to win a market on two occasions; FPGAs (Field Programmable Gate Array). Intel took two shots at the FPGA business and failed both times. They failed not because they couldn't do FPGAs, they stopped the offensive after learning that the total FPGA business was well less than a billion dollars and very high support.
So, Intel wins and loses. However, if the market is large enough and the business can benefit from unique Intel technology, Intel will win it. The path to dominance is sometimes messy, often very expensive, it might includes some embarrassing failures, but in the end, Intel gets what Intel wants.
Intel is on the offensive again in the device mobile chip business. Again the battle is noisy, expensive, and messy, and Intel has and will likely continue to make mistakes. Again the competition is very capable; Qualcomm (QCOM), Samsung (OTC:SSNLF), Taiwan Semiconductor Manufacturing Company (TSM), and even frenemy, Apple (AAPL).
In the end, Intel will bring all the pieces together in a fully integrated family of device mobile SoCs that will be built on processes that will make the chip at one fourth the size and cost of the nearest competitor. Intel will price the products at the competitor's cost and still make 65% gross margins.
The Apple A7 chip is made by Samsung on a 28nm process. TSMC was rumored to be in line for part of the business. They didn't get it. On the latest TSMC Conference Call they cited soft 28nm demand for a soft fourth quarter. Obviously they had the capacity to get some part of the Apple business. Could it be that the apparent leader of the foundry 28nm business is having problems meeting Apple specs or, heaven forbid, yield?
What if TSMC and Samsung can NEVER get past 28nm? Even TSMC Chairman Morris Chang expressed concern in public that if TSMC couldn't keep up with Intel the consequences are "too horrible to imagine."
Intel is shipping Haswell processors today made on the second generation TriGate process while TSMC hopes to ship 20nm planer products, if you listen to the CC, sometime between…umm tomorrow and the middle of 2015 and it is NOT clear exactly when. There is major discussion in the technical community about whether or not the TSMC, Global Foundries, and Samsung are selling smoke or substance on sub 28nm processes. With the possible exception of Samsung, I have a lunch bet with friends (big spender, me) that the others NEVER ship anything of commercial value below 28nm.
What is truly amazing about Intel is the company's tolerance for risk. Not many recognize the fact that, with each tick or tock of the Intel cadence the company is betting their very existence on the success of a new process or CPU architecture and have been doing this nerve wracking process since the days of the 80286 in the 1980s.
For example, right now the company is shipping only Haswells; a year ago Intel was shipping only Ivy Bridges. This $50 billion company puts itself on the line every year. With this in mind, the noise about "missing mobile" is simply an annoyance.
Intel shareholders have to remember that Intel is on the offensive; all the other device mobile chip suppliers and foundries are on the defensive. That is not a comfortable place for them to be. Intel will win.
Intel is a buy and 2014 is the year.