So if that number the article referred to 50% of output shifted to Rambus, then roughly 15M 64M pc100 parts would be converted to Rambus.
IF you look at most of the predictions of mix this year, the market is going to continue using tons of 64M PC100 devices and taking 15M per month out of the output WW would cause one heck of a short supply.
The other thing I read that could also impact the market, if the
current yields are 80%(lets say) and suppliers convert over to RDRAM
that portion of wafers are not going to yield the same as 64M sync, thus losing some % of output in 99. Some of you have a better handle on the numbers, and could provide a better analysis.
The last thing, that also seems concerning, is the balance of planned output of RDRAM vs Sync. IF a supplier loads the FABs incorrectly to the demand, they could end up with too few of one device and too many of another.I know it hard to imagine having a warehouse full of RDRAM and a shortage of PC100 sync, but I think with the variables and uncertainties that are currently present this could happen to more than one supplier, and depending on one of the variables could happen to all the suppliers and we could very well have one heck of an imbalance.
I am not predicting anything, but it is something to thing about.
I would like to know if anyone thinks these things could be viable concerns, and any other factors that I missed.
Did everyone like the swings this week with MU. Thats the way
I like a stock to be, I wish they had more conferences.!
Good trading. |
