Cypress has a history with Apple, supplying parts for the iPod. And Marvell now owns Intel's (NASDAQ:INTC) old mobile division, which is a potential microprocessor supplier.
If Cypress has a piece of the iPhone, it would help boost the company's PSOC business that has been right around the corner for too many quarters. For Marvell, it could be a small boost for its cash acquisition that Wall Street punished the company for by hammering shares upon the Intel deal's announcement last year.
Cypress is a compelling play for other reasons as well. The company has a 65% stake in SunPower, a fair amount of net operating losses, what I expect will be improving gross and operating margins in the coming periods due to product mix, and a likely debt repurchase that will remove some overhang on shares.
Also, Chapman Capital, an activist fund with a great reputation of success, now holds about 1.5% of Cypress' outstanding stock. Chapman is making a push for Cypress to consider going private, and pushing for the company to further monetize its stake in SunPower.
More recently, Cypress has announced plans to cut back on Capex, move away from Moore's Law with its "No more Moore" strategy, and competitors in a key end market have shuttered operations leaving Cypress as the only viable option for companies.
I believe the semi-industry is within a few quarters of clearing up some inventory issues, and similar to the last time we had oversupply in semi's in 2004 the group should soon find traction and rally ahead of what should be a solid second half of the year for orders. Cypress is somewhat of a special situation play in the semi space with the activist component, making its already compelling valuation discount to peers after stripping out its stake in SunPower from the valuation even more attractive.
For Marvell, I believe the company remains plagued by a historical propensity to underperform. Even so, Marvell could see a boost in demand for its WLAN products with 802.11n likely ratified by 2008 as a standard. Also, the HDD business, which was hurt by a slowdown in PC demand ahead of the Vista launch and a key customer being acquired by Seagate, should benefit from a cyclical boost to PC demand in 2007.
I believe the general estimate for 7% PC unit growth in 2007 will prove conservative. This is based on some backwards math to figure out how most analysts got this number. I believe the replacement rate of PC's moves from 1.5 months at present to as low as 1.3 months. Given that 2/3's of all PC's sold are replacement units, the 2 month shift, which will be driven by the Vista upgrade, is strong enough to get me to 10% to 14% unit growth for the 12 months ending June 2008.