After selling their X-Scale processor division to Marvell (MRVL), Intel (INTC) still seems interested in the smart-phone, MID market - which they hope to conquer using their Atom series of processors. While these compare favorably in computing power with the TI OMAP 3440, I think that INTC once again has fallen into the legacy trap. The Atom has a whopping 512K on on-board 8 way L2 cache - which will make it worthy of running legacy Microsoft Applications. While it's sure in Intel-Lite, I do not see it as competing with ARM’s (ARMH) Cortex-8 ARM11 architecture [on which the OMAP 3440 is based].
ARM can do well in the MID market, and has both TI (TI) and nVidia (NVDA) working on making innovative small-footprint, low-power chips, and can extend its lead through their leadership in the mobile internet market - where there is at least one ARM based device in 90% of all mobile phones. Looking at the OMAP 3440, it can handle HD Video up to 720p and can interface with a multitude of communication platforms - including USB, UART, and SDIO links to Bluetooth, mWLAN, etc.
ARM’s financials are pristine, and their latest revenue and profit numbers have been hurt by the strong UK pound. But since they are fabless, and essentially, the most successful IP company, they have gross margins in excess of 89% and I like that number [the CME (CME), Mastercard (MA) and Visa (V) do better]. In fact, they returned almost 90% of their cash flow as dividends and stock-buybacks [to stockholders] in the last twelve months. They have 1.27 billion shares out [excluding treasury stock] - that translates to 430 Million ADRs - which makes ARM fairly valued at the current quote of $6.25 per ADR, and actually a little pricey on a PE basis.
I’ve done a significant amount of research in this technology, but am searching for ideas for writing useful, thought provoking articles. All ideas welcomed to bapcha@bapcha.com. We are still looking at a not so good six months for semiconductor companies, where I think a recovery will probably start in late 2009 - as the leading indicator [capital equipment companies like Applied Materials (AMAT), Novellus (NVLS), Lam Research (LRCX)] is not yet signaling a recovery in the market.
MIPS (MIPS) made a terrible mistake in buying ChipIdea. They spent in excess of $145 million in cash to buy the Spanish provider of physical IP in an attempt to emulate ARM, but there is one significant difference between ARM’s acquisition of Artisan vs. MIPS’s acquisition of ChipIdea. Artisan was profitable and still is generating record Licensing and Royalties revenues as a division of ARM. ChipIdea [I am assuming that MIPS's pro-forma numbers filed at the SEC are correct] had gross margins of about 30% - while the rest of MIPS generated gross margins above 90%. Sure, ChipIdea grew from zero to $38 million in six years, but most of what ChipIdea does is done in-house in the bigger semiconductor companies like Broadcom (BRCM).
Anyway, my pick for the winner of the MID market is ARM. I expect significantly more interest in the stock when the first phones that are based on Google’s (GOOG) Andriod are released, since they will all have an ARM CPU powering them - much like 80% of the mobile market today.
Disclosures: Long Intel. No positions in other companies.
This article has 2 comments:
Chipidea is almost 10 years old, not 6.
Chipidea's revenue in 2007 was way above $38 million.
But in the 21st century's age of information, and given the fact that Chipidea is PORTUGUESE, calling it Spanish is just plain dumbness.
Geezzz, at least google your data, Bapcha!