AMD Divests TV Unit to Broadcom

Today Broadcom (BRCM) announced it would buy AMD's - consumer television chip division for $192.8 million. AMD had announced that this division was up for sale, and was being treated as a discontinued operation, at its July 17, 2008, Q2 AMD analyst conference (summary).

This segment was acquired when AMD bought ATI in late 2006. It is difficult to make direct value comparisons because this particular line has not been singled out by AMD or ATI. For the quarter ending May 31, 2006 ATI reported its consumer segment revenues, which included more than the TV chips, were $143 million, but it also had quarter revenues of $652 million and was heading into a steep overall revenue decline even before AMD acquired it. By the time the acquisition was complete, reporting for Q3 2007 (see AMD analyst conference summary for Q3 2007), AMD was reporting consumer revenue of only $97 million, and that also included more products than the television chips.

Broadcom, by the way, provides chipsets for AMD's Opteron server motherboards, so they have a prior relationship.

AMD paid almost $5 billion for ATI, a major overpayment in retrospect, since NVIDIA (NVDA) was about to capture much of ATI's business. AMD desperately needs cash as it continues it's David and Goliath game with Intel. But it is important to note that at one point AMD management was calling video chips for large-screen TVs a growth area.

You hear that sort of thing all the time at technology companies: our sorry asses will be saved because we are into the next big thing. Forget that we failed at the last big thing. How many technology companies were going to make a mint for investors by plunging into wireless Internet (WiFi & etc.) back in 2004? How many actually did? As usual, the companies that won the last round, like Cisco, mostly won the new round.

That said, I think AMD is wise to focus on its core competency, computation, including graphics computation. It wisely shed its memory chip division before it picked up ATI's graphics expertise. Some things are going right for AMD now: quad-core Opterons are performing impressively, Intel's upcoming Nehalem chips don't appear to be anything great, and the ability to have high-end graphics and high-end CPUs is pulling customers away from both Intel and NVIDIA. On the other hand AMD's stock price is in the swamp, and both Intel and NVIDIA have been happy to compete with price drops when they could not compete with AMD's advanced technologies. AMD needs to start generating cash from operations in Q3; the cash infusion from the Broadcom deal and last quarter's sale of fabrication equipment is needed just to help balance the debt from the ATI acquisition. If AMD can't sell enough chips at a profit, Intel wins by default no matter. Intel's legal problems won't come into play in a meaningful way until 2010 or even 2011. Even then, paying a few billion dollars to AMD will be worth it to Intel; it will have been a small price to pay for stopping AMD in its tracks in 2004-2005, when ordinary competion could have led to Intel's overthrow.

AMD is a very risky stock, but I like the fighting spirit of the AMD team, believe the stock is currently undervalued, and own the stock.

But keep diversified.

More data:

www.amd.com
My AMD main page
www.broadcom.com
www.intel.com
My Intel main page
www.nvidia.com
My NVIDIA main page