So, PF did not say what is the total inventory, but start to cut back in early Jan of 98, however based on the information it is about 0.2MM units. So, you tell me how long will take CPQ to clean those inventory with a total sales of about 13MM units per year ? and I do not think the inventory alone of CPQ will cause a 10% shrink of revenue of INTEL, am I right ? It must be something wrong, the competion and the BTO business models which force INTEL to place huge inventory inorder to meet the on time delivery requirement from the
box makers. I think you realize that INTEL also sit with big inventory too !!!!
Yes, AMD had a yield problem , but IBM do not. The yield problem can be caused by either the process itself or the design, I think you realize that. Will IBM's process fix the yield problem ? I think so, I think you realize that before switch fab , you generally had to qual the fab to make sure the yield , quality is ok, am I right?
As to the National's deal with TSMC, it is not a rumor, and had been confirmed locally in Taiwan.The newspaper already talked it.
The INTEL's myth is "INTEL INSIDE", but it did not play now when people realize that CPU from INTEL or CYRX or AMD really make no difference, the difference is the price.
This is the competion from the low end, and in the middle end, not too much application for middle end Pcs at all, then why should I bother to spend more money to buy a middle or even high end PCS? especially , most of my application on PCs are word processing and C programming ? That is another problem for INTEL.
INTEL has been released new CPUs too fast , which make consumers or customers hesitate to
buy or upgrade. Why should I upgrade now when I know that my investment will be obsolete pretty soon ? The new BTO business model implemented by most box makers will force
INTEL to build huge inventory to meet the on time delivery standard.
I jump directly from 386 to pantium, and never think about upgrade to PII at all. My current note book meet my needs pretty well !!!! I think many people have the same thinking !!!!! This is the problem for INTEL, so you started to see the impact this quarter, and will see more in the quarters to come. If INTEL can still maintain a 70% of the market share of
CPU, three years down the road, then it is very lucky. The PC market will keep growing fast
, especially as the sub$1000 step in, many people can now afford. But that does not necessary benefit INTEL, since people now have more choice of CPUs, and the competition on CPU price will be fierce, this will of course benefit all those box makers. We already see IBM, CPQ endorse AMD and CYRX.
As to the high end, you know what will compete with Merced. You do not need me to mention it over here. To tape out at the end of 1998, probably will take at least 6 months
to go to formal production if you are very lucky for such complex chip, and the system based on Merced won't be available at the end
of 1999. Of course there are many uncertainity down the road about Merced, such as:
yield, debug ....etc, you name it . |
