Trey, I don't believe that the Koreans are making money on 64MB chips. Most customers are indifferent between 16MB or 64MB chips - all they care about is price per bit. When you buy a SIMM, do you ask whether it will have 2 chips or 8? Thus the price of 16MB chips is a limiting factor on what a competitor can charge for a 64MB chip (and vice versa).
MU is perfectly capable of making 64MB chips today, should they choose to. They have the tools to do so, and all the necessary technolgy. In fact, you can be sure that they have already made some. They will make the transition to 64MB chips when the 64MB chips become more profitable for them than 16MB chips, just as they made the switch to 16MB chips when 16MB chips became more profitable to them than 4MB.
Remember though, that the lower their costs are for a particular chip, the longer that chip remains profitable, and the later they will make the transition to the next generation. Since MU is the (a) low cost producer, it isn't surprising that they are able to remain in a lower generation for longer than their competition. Thus I re-assert that they drove the competition to the next generation. The competition had higher costs at the prior generation, so they were forced to make the transition at an earlier point.
As for the profitability of 64MB chips, remember that initial yields will be low and costs high. Yet because prices keep falling, they are unable to achieve the situation you describe for them, i.e. reaping monopoly profits for being the initial producer. No, I think the Koreans are bleeding just as much red ink at 64MB as they did at 16mb. And I think you will be disappointed by how smoothly MU manages the transition to 64MB when it deems that the time has come.
Yep, in making memory, cost is king.
good luck,
Carl |
