Semiconductor industry will go through a major transformation in the next ten years. The foundries will be the biggest beneficiaries of this transformation and the fabless comopanies will be the net losers.
I found #1 and #2 interesting. The market does mature but because the industry is continually evolving, "maturity" changes and we see the cycles again. Also, the boom-bust cyclicality is at least partially dependent on the worlds economies which are often governed by factors outside the control and outside the predictions of those within the industry.
Malcolm Penn has hit the nail on the head with most of these, but I take exception with #5. For what it's worth, here's my 2 cents worth of further elaboration:
Myth 1: Not only are the cycles not over, the last one had the largest amplitude swings ever. The semi business is behaving like an unstable feedback system. Like any other feedback system, if we don't dampen those oscillations, something is going to break.
Myth 2: A corollary to Myth 1. When the market grows, we have a boom phase. When the market declines, we have a bust phase. No maturity there.
Myth 3: ASPs, like anything else in economics, reflect the supply & demand intersection. Perhaps when supply is large & demand is weak, cash-rich companies can wrap dollar bills around their ICs to maintain market share, but that is not sustainable. Our industry thrives on cost reduction, but not so much on gross margin reduction. Nobody works for free, and if the stockholders can get a better return buying T-bills instead of building semiconductors, then they will buy T-bills.
Myth 4: Fabs definitely have strategic value for those who are not such large foundry customers that they can dictate terms. Just ask any fabless company that is on allocation and crying about how big their sales could be if only they could get product so they could ship it to starving customers.
Myth 5: The fabless model IS a success if you are one of those large foundry customers that is up near the front of the line. Broadcom is fabless. Anybody want to claim they are not successful?
Myth 6: This might appear to be a conservative approach, but it's actually quite the opposite -- like going to the roulette wheel and instead of placing multiple bets, the company puts all it's money on lucky number 13!
I love point 6. I think we are at a point where we could likely benefit more from having a more diverse process landscape. I think that many of the post CMOS technologies could be made to have their place in industry assuming we can get them to work, and the only way we can get them to work is with large scale industrial R&D.
Interesting take...I agree stronlgy with #5, in fact I am surprised this is listed as myth, I thought this is widely accepted that a fabless chip business is very tough and not many VCs would touch an investment in one, simply no ROI...Kris
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...