Shane,
More thoughts about system-on-a-chip and other larger companies which may want to get involved. I thought this mention from Michael's notes of the shareholder meeting was interesting:
He made mention of LSI's reliability for timely implementation. Citing many former DRAM semi's in Asia that tried to reorient to SOC product strategies, some LSI customers were attracted to the cheap prices they were being quoted for asian implementation. He said, "Give them a try". They did, and did not get their products out the door thus missing their "window of opportunity". They have come back to LSI.
I am sure that the DRAM makers and others would love to start competing in LSI's ASIC and eventual SOC market. But it's not something you can just pick up and do. You have to hire designers with the right expertise, you have to have the right EDA tools, and you have to optimize your fab processes. If you are going to change from DRAM to ASICs, you also have to change your organizational mindset from a commodity product to application-specific and customer-specific designs. Sure it can be done, and they may have high motivation to do so, but it will take at least a couple years and not all who try it will be successful.
Other companies like Intel may have a better chance, but it would not be a simple change for them either. INTC has done pretty well in markets other than microprocessors, but they are markets that require intimate knowledge of the microprocessor (motherboards, core logic) and therefore INTC started out with a big edge. Their move into graphics chips has not been as dominant, and even there they had to make an acquisition. They may eventually become dominant in graphics, too, but it isn't guaranteed, and the competition looks a lot stronger than INTC at this point. So I'm not too worried that they can invade LSI's ASIC markets at will.
G.P. |
