control their own future...fab or fabless?
contract fabs that do not have their own end products, but manufacture for others, have by necessity lower margins than the fabless companies they serve, and furthermore, they do not control their own future in the same way that companies serving end demand do.
Besides the fact that Tower does have their own end-products, and specialty technologies in hot emerging markets (i.e., non-volatile memories and CMOS image sensors), one must consider that the fabless player is not necessarily the preferred stock investment for all market conditions and situations. As the following article states, "Large fabless and larger IDMs have made moves to lock down capacity, and I understand smaller buyers of wafers may have some difficulty getting access," :
siliconinvestor.com
I've wondered whether a fabless company is concerned about losing a grip on their IP once they farm out their fab business to the Far East. But what choices do they have? Maybe they figure that if they keep improving their designs every six months, then they can worry less about having to compete against their misappropriated, 'antiquated' technology? Pity the graphics chip players (SIII, TDFX, etc.) caught in this viscous cycle, eh Zeev?
Nonetheless, I can forsee certain benefits to Israeli fabless companies that have a state-of-the-art fab at their disposal within their own country, would you not agree?
Regards,
thecalculator |
