Intel (NASDAQ:INTC) is the largest semiconductor company as measured by dollar volume in the world. The company is embedded into every aspect of computing. Since, for various reasons, PC sales have gone flat, many industry pundits are having fun declaring the "Death of the PC", not the "Improvement" or "Change" of the PC; these folks are forecasting the Death of the PC. They might as well forecast the death of human civilization.
The semiconductor industry and computing industry that it has spawned is all about change and nothing about death. If the miracle of making computers out of a few hundred "light bulbs" called vacuum tubes 60 years ago to making computers out of billions of transistors fabricated on a silicon chip the size of your thumb nail today doesn't cause you to catch your breath, you have no soul, and should try to find your way back to the caves that we all came from.
So there, now you know that I have a high regard for human intelligence and the constant changing creativity of the electronics industry as enabled by the semiconductor segment of that industry.
Back to Intel. The Intel outlook for 2013 indicated a slow start for the first two quarters with a return to growth in the last half of the year. So, what will we see that causes this change only a couple of months from now?
First, we're in a "Tock" cycle. What?? OK, OK, let's talk about Intel's Tick-Tock strategy to introduction of new microprocessors. Tick-Tock is all about making one step at a time in a very complex process in order to make sure that things don't get completely "wrapped around the axle". The Tick part of the cycle is when Intel takes existing microprocessor architecture and simply moves it to a smaller manufacturing process. Since that device has been built in large volume on the existing process, if there are problems with the new process, Intel knows that the process is the problem, not the architecture. That basically means that each new architecture is built on two process nodes.
When the new process and architecture is stable, Intel begins to build a new architecture microprocessor on the now proven smaller process node. That is the Tock part of the cycle. For example, Sandy Bridge was the last "Tock", and the current Ivy Bridge is a "Tick".
The Sandy Bridge (Tock) was introduced in early 2011, the Ivy Bridge (Tick) was introduced in April 2012 on the new 22nm process. The Haswell (Tock) is being ramped up about 12 months after the Ivy Bridge, so the cadence has quickened. The Sandy Bridge and Ivy Bridge were essentially the same architecture with no ultra-heavy emphasis on power reduction. Both parts relied on power control on the mother board.
Haswell, being a Tock, is a horse of an entirely different hue. Power reduction on the Haswell is the prime directive. Idle power is 1/20 of the predecessors. Since, in normal use, a PC is in idle mode 95-99% of the time, the battery life on Haswell PCs will be dramatically better than previous versions. The Haswell does this through the use of a "Voltage Regulator". Calling this chip simply a voltage regulator is like calling a Masaratti a grocery getter. The chip is situated right next to the Haswell CPU, in the CPU package. This chip is capable of controlling power to the Haswell CPU at up to 320 different points, if necessary. This is truly an unprecedented level of power control for any computing chip. At idle Haswell will consume as little as 100mw.
Haswell could very well be the catalyst that excites PC users into upgrading, particularly since the Tick of Haswell is just a shrink to 14nm.
The Intel reference design called the North Cape is claiming 13 hours of battery life as a clamshell PC and 10 hours as a tablet. I seriously want one of those today!
Some PC OEMs may choose to take the power saving regimen a step further with local screen refresh techniques.
One of two major impediments to progress in the mobile area has been the lack of a multimode, multiband, voice and data 4G LTE chip. This is the chip that is required in the U.S. It must be able to work in different frequency bands and be able to back down to 3G and 2G where the 4G infrastructure in not yet in place.
Intel is due to introduce a full function LTE chip "in the first half". The chip has existed in pre-production version for a while and has been tested against other new LTE chips coming to market. Actually the testing rated the Intel chip number one among all tested.
The part number will be the 7160, so when you hear that part number announced, you have my permission to party.
Don't forget that a little further out we have digital radio coming. Intel is well down the road on that technology.
We have been in a memory crisis and came close to near domination of memory by Samsung (OTC:SSNLF) and Hynix (OTC:HXSCF). That issue seems to have been resolved. I still think Intel/Micron (NASDAQ:MU) could get to 450mm wafers before anyone else and sooner than anyone else. 450mm wafers make NAND flash memory really cheap. Think 2015 for 450mm.
With Haswell already shipping and the LTE and memory impediments gone, Intel can: 1) Enable unique convertible Ultrabook formats, 2) Get involved with making Apple (NASDAQ:AAPL) "A" chips, if that is the plan, 3) Provide smartphones and 4G connected tablet and smartphone application processors/LTE to American carriers.
Intel could be the best investment for the next several years.
Disclosure: I am long INTC. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.