My 1st Million At 33 - yes, you can do it too

I’m starting a Future Trend series to express my views for the future technological trends, and how it may relate to your investment. This is the very first one. I hope you will enjoy it. Please let me know what you think via comments.

With Moore’s law of doubling transistor counts every 18 to 24 months, integrated circuits are facing an unprecedent level of integration and increase in both processing and communication power that is not possible before (or just 24 months ago). While in the past, the Moore’s law has primarily benefited consumers in terms of more speedy processors in the computer, the current trend is now onto multi-processor architecture simply because of two primary reasons:

1. Power density as the processing clock frequency increases is posing a seirous technological challenge.
2. Everyday software applications cannot benefit from higher speed, but rather from multi-processors (especially with dual processors).

With Intel losing market shares to AMD in the last two years, partly because Intel was late both in 64-bit desktop/server processor, the CPU market is undergoing a commoditization process in which price, not brand, is what matters more. Being wary of AMD competition, Intel has divested its communication interest, and refocused its effort on processors. Intel has beaten AMD this time around in introducing quad core processors, shortly arriving in 2006.

However, doubling or quadrupling the same CPU may not be doing much for consumers. With so many transistors increasing exponentially on a single chip, the amount of IP (intellectual property) that can be designed by engineers can barely keep up. Instead of doubling or quadrupling the same thing over and over again, the more important industry trend is towards more integration of different functions for the goal of reducing overall bill of material (BOM) cost. After integration, a single chip performs the same amount of functions which were achieved by 2 or more chips, and using less PCB (printed circuit board) area, and less overall cost. In the endless drive towards lower cost, certainly the company that is able to deliver the highest degree of integration would win at the end. IC design companies on this caliber are few. I would only mention BRCM, MRVL who are eating or ate away market shares once belonged to INTC, STM, CNXT (for disclosure purpose, I own/owned at least one of the previous five securities at different times). Both companies however are under stock option back-dating investigation and are delinquent in their SEC filing.

The integration of different chips is allowing box/device or hardware makers to provide many more functions at a lower cost and shrinking geometry. The best example is the current available cell phones, many of which you can use as MP3 players, digital camera, browsing web, and checking emails for cellphone PDAs. The low cost examples are DVD-players, DVD-recorders for PVR. I’ve seen rock bottom prices under $30 for DVD players, and I bought my current Lite-On DVD recorder at Costco for only $90. It replaced my VCR, and I used my DVD recorder for archiving my mini-DV tapes.

As IC companies cannabalizing each other’s market share and overall revenues, the survivors may be very few. Despite all these, IC revenues each year still increases, due to newer applications and higher market penetration. However, gradually the picture for the losers won’t be pretty. If you do own individual companies in this space, I advise you to pick your stock carefully. Make sure the company that you picked is on the right trend of more integration of mixed signal (analog & digital) chips. If you are not so sure, you should own sector ETF such as SMH instead. If you are confident about which companies are good and which ones are bad, you could consider hedge your own bet with partial short position in the worse companies.

More comments on INTC vs AMD. If you look at what Intel (INTC) has done, divesting communication business, while AMD acquiring ATYT for graphics business, you should realize that INTC is really on a dangerous track especially knowing that integration is the game in IC. Given that INTC graphics chip has always been at the low end, AMD acquiring ATYT may become the next biggest strategic threat to INTC. Another good reason to buy SMH, instead of individual stocks.

As they say, QQQQ leads the general market, while SMH leads QQQQ. It is important to watch SMH simply because IC semiconductor companies are usually the leaders of the technology.