AMD 2006 Analyst Day
January 4, 2007
2007 will be interesting. AMD expects to drive microprocessor growth at twice the rate of the industry. Source: 2006 Analyst Day (p. 34):
One of the things that was interesting about the Analyst Day was the way in which it impacted the financial community. AMD's stock rose by more than 10%.
At the very least, the conference convinced the financial community that the acquisition of ATI is going well.
AMD said alot of things that were true at the Analyst Day. 2006 was a phenomenal year.
However, in the last quarter AMD's profit margins were hurt. Margin is about the most important financial metric for measuring IT companies. This could bode ill for AMD.
AMD is attempting to change the industry in a big way and take the industry on a different course.
Do not look to AMD to push the envelope of performance. Instead the company shall put more emphasis on leveraging existing technology.
One way to increase performance is to double the number of cores on a CPU. AMD has insisted, however, that it wants "No Core Wars" (p. 9):
Similarly, one of the biggest bottlenecks holding back performance is the memory subsystem of computers. Yet AMD intends to stick with DDR2 for the time being.
Rather than a gigahertz wall, or a wall of transistor size, AMD speaks of a diversity wall. One size no longer fits all. All too true. The era of the general purpose CPU is at an end.
AMD intends to provide a platform for specialization. The platform is to be modular, flexible, and open to other industry participants.
AMD shall go after the mainstream and niche markets within the mainstream.
At the same time, AMD is hedging its bets.
AMD didn't buy ATI solely so they could open up AMD's systems to others. They bought ATI so they could--among other things--have a chipset business so they could better tap into the (lucrative) mobile market.
The strategic, long-term objective of acquiring ATI is to sell more microprocessors.
Graphics chips and chipsets have lower profit margins than microprocessors. The only way that the acquisition of ATI makes sense is for AMD to eventually sell more high margin CPUs. Otherwise they are simply lowering their margins, which will hurt their stock.
Do not expect AMD to be successful overnight. AMD does not anticipate the investment of ATI to really start paying off until 2008. So in spite of AMD's optimism, 2007 may be a year of pain for AMD.
AMD is counting on support from the big players in the industry--HP, Sun, IBM, Dell--on grounds that the industry likes choice.
AMD has a clear advantage in graphics over Intel. It needs to leverage this advantage in computer graphics and in CE (consumer electronics).
Possibly AMD should start taking seriously the race for more cores. As long as there are applications out there that take advantage of multi-chip packages (MCP), AMD may have trouble convincing the industry that doubling the number of cores is not a good thing.
In retrospect, one can only sit back and wonder what would have happened had AMD been able to acquire NVIDIA, instead of ATI. NVIDIA and AMD would have made an awesome combination.
If things go south, maybe in time NVIDIA will acquire AMD?