AMD's Troubles and Why Qualcomm Should Buy

As ente

rtaining as these rumors are and as easy they are to dismiss, there is a serious note to such stories. AMD has attractive assets, while it isn't exactly competitive at this time anymore. A deep-pocketed buyer with a complimentary set of assets could be very interested in AMD.

There is an interesting discussion in Silicon Valley among analysts that suggest that an acquisition of AMD isn't the stretch we believe it to be. Dell buying AMD would be big surprise, especially in the background that the company's core business are PCs and an acquisition of AMD would put Dell in the line of fire of Intel. As dependent as Dell is on Intel, that would be suicide. If there is a PC vendor that could be interested in buying AMD, it would be Apple -- if the company was to build its own CPUs, but that does not seem to make any economic sense for Apple.

The options

Here is another idea: What if there was an ARM vendor that wants to break out of the old? ARM itself is unlikely to leave its successful business model, but actual chip makers could be using AMD to protect themselves from the coming Windows on ARM era. There is a lot of knowledge in AMD and x86 may be an interesting diversification.

A quick look into the balance sheets of ARM licensees reveals that there are just two options and two companies that could afford to buy AMD and sustain a war with Intel. Samsung and Qualcomm. TI or Freescale may be interested as well, but their financial covers are too thin to swallow a company that could cost somewhere in the range of $10 billion. Qualcomm has about $10.5 billion in cash and Samsung has a bit more than $11 billion available in cash.

The chip manufacturing horse race is already led by Samsung and Intel as the two dominant semiconductor companies worldwide. Samsung and Intel are currently considered to be the only manufacturers that can drive chip manufacturing beyond the 20 nm node and they are believed to have enough financial resources to build the $6 billion+ fabs that are required to make such chips economically feasible. However, Qualcomm appears to be an even more interesting buyer.

Qualcomm acquired AMD's mobile graphics business a little over two years ago. The Imageon processor has since then dissolved into Qualcomm's mobile GPU business and the Adreno GPU series. There is already some business synergy.

Within 18 months, Qualcomm as well as other traditional ARM vendors such as TI will get a new opportunity with tremendous competition in the Windows ARM segment especially from Nvidia, which is putting out stakes for the lead in this market. As inexperienced as Nvidia may be in this field, as serious is the threat to ARM vendors: Nvidia will be using its chip building expertise and GPU know-how to become the leading ARM vendor for Windows systems. AMD could deliver invaluable technology for Qualcomm and provide business options to boost its ARM business and explore the x86 market. It would be a substantial risk for Qualcomm, but the company would not burn any bridges with Intel and could supply the funding for a more competitive AMD that would have to get much more nimble, however.

The AMD Perspective

On AMD's side, we know that CEO Dirk Meyer had to go because he did not react quickly enough to cater products to the tablet and smartphone market. While AMD maintains that there will be tablet processors available this year, the company has empty hands in a buzzing market right now -- a market that holds the promises of the biggest opportunities for processor makers in decades. Next to Meyer, COO Bob Rivet was also let go. There may be other reasons, but there is a slight chance that the company is being prepared for a substantial shift in product strategy.

What made me question the firing of Meyer was the fact that there aren't many executives that can successfully lead a company like AMD -- people that can sense trends and motivate employees with the necessary vision, talent and charisma. If you will, the AMD job with all its economic ups and downs and the continuous threat of being extinguished by Intel isn't exactly an any-type-CEO job. Especially in this critical stage for AMD, it may have been more effective if chairman of the board Bruce Claflin would have stepped on Meyer's toes to change the direction of the company. Claflin must already have a candidate in mind -- someone like EMC's Pat Gelsinger, who led the engineering of Intel's 486 and Pentium processor -- or he is simply cleaning out the company so it can be acquired.

Either way, we will see a pretty significant development for AMD soon. If Claflin fired Meyer without backup, the effects for AMD will be equally significant, albeit in a negative way: AMD is currently doing well financially and its CPU lineup looks promising, with the exception of the gap in the mobile range. However, we also know that it is part of AMD's culture to shoot itself in the foot when the company is doing well. Self-inflicted vulnerability has always been a part of AMD's history. Remember how dominant AMD was with Opteron and the Athlon X2 in 2004/2005? Back then, it was pure arrogance that punished AMD and allowed Intel to take back the technology leadership.

The statement that AMD will be acquired is as much speculation as it can be, but Qualcomm seems to be a company that could be very interested in doing just that. Are we seeing the final chapter of AMD as an independent company right now?