Here's a great idea for a business model. We'll take on the best semi manufacturer in the world. His process technology and manufacturing smarts are beyond question. He has a solid marketing approach and has divided the market into several segments with strength bordering on the obscene in several of the higher margin segments. Speaking of margin, he is running in the high 50s. He is raising the bar dramatically on operating frequency. He is simultaneously dropping his costs per unit substantially. There are occaisional marketing screwups like the floating point snafu and the pcid public relations problem, but on balance a superbly run company hitting on all cylinders.
I will take them on, though my finances are shaky, my processes wilt under stress, my marketing is so-so and my investor relations have been a disaster. I am counting on my competitor not responding when I take significant low-end share. I will take a high risk approach to improving my manufacturing picture by using a brand new process at a brand new feature size with a brand new design. I will bet the farm.
My son is a newly minted MBA from an ivy league school and while he was in school he mentioned in passing that, in todays world, it is suicidal to compete solely on price. We weren't talking about microprocessor manufacturers but he told me that, in a high tech world, where there is or can be differentiation by features and rapid changes in product offerings, relying solely on low price ain't gonna cut it cause your competitors aren't sleeping.
If AMD can survive at all, it will be as a division of someone else, not as an independent company. They don't have the depth of pocket, the business model, the technology edge or the management depth to play in the league they have chosen to play in. In a world where perfect execution of their game plan could still fail, imperfect execution of their game plan dooms them. And they have shown an alarming propensity to stumble.
Smaller, nimbler companies can beat established giants. But it's a high risk game and you have to have an edge. And near perfect execution. It helps if the giant goes to sleep, like IBM in the 80s.
Regards,
Burt |
