He has a very valid message mostly because of Intel's double duty as a fab and design company. They have a massive amount of capex and foundry costs that must be paid for. AMD only has to pay for R&D, they have muuuuch less infrastructure weighing them down. If the desktop market shrinks and Intel doesn't change their attack plan, they are left with fabs which have more or less static costs, and profits which are smaller. Without large volumes and high margins, Intel could have a tough time with all of their foundry "baggage". Just something to consider, as Intel is more vulnerable to weaker profits in comparison to AMD at the moment I would say. AMD can lay off employees, close an office maybe, Intel would possibly have to close an extremely expensive investment in a fab, or possibly start contracting fab work out (and lose their competitive advantage).
None of this makes any sense.
A. In the plunging PC market of Q1 2013, intels profits shrunk from $3B, to $2B. AMD lost $100m. When you are making $2B per quarter profit, it's a long, long way to losing $100m.
B. intel has 86% of the PC market, and a large advantage in process and performance. They also have many billions in the bank. If threatened by "dropping profits", they can lower prices until AMD can't compete at all anymore. Then they could raise them to whatever they want.
C. intel is among the world leaders of fabbing. Anyone else doing a chip as complex as a 4770K at 22nm? Nope. Every silicon design house in the world would love to have intel cooking chips for them. That is a bigger deal than "we're selling 4m low end APUs to the console guys a year".
D. AMD has already laid off tens of thousands, sold their buildings, sold their fab, etc etc etc.. You get to a point where you need to keep some people around to do the work. An ex-AMD engineer has already told us Bulldozer turned out worse than expected because there were far too few human engineers, and it relied far too heavily on computers for design. (which really annoyed me, I buy AMD when I can because of point B)
E. intel already has changed their attack plan. Look at what they're doing with graphics, power, IPC, and diversification.
That guy Roy's analogy just doesn't fly in my opinion. The desktop market has slowed down a lot because CPUs have been good enough for 99% of office apps since C2D days so Ma and Pa don't need to upgrade anymore and haven't for a long time.
However; businesses are on 2-5 year refresh cycles and they buy 86% intel, and I don't see that switching.
Facebook
Twitter