Steve and Dave,
Reading both of your posts is like watching a fencing match, one
thrusts and the other paries and then takes the offense. Good
reparte' but I declare it a draw.
You both have valid points to your argument but I feel I must side
with Dave's arguement. Investment money is tight, risk investors
(banks/corporations) are unwilling to invest in short/mid term
Korean chip manufacturers because they are over-extended and their
income is minimal or non-existent, at this time, so it is a bad
investment risk in an oversupplied and weak DRAM market. Fabs have
an option to upgrade existing equipment and processes to get them
to the newest 0.25uM device geometries at a price cheaper than
building or purchasing the newest equipment at this time. This will
get them over the short to mid term technology requirements and
allow them to survive, but with tighter belts and with reduced
expenses.
Another way is to diversify the product portfolio and produce
ASICs or other foundry chips, where the market is not so soft.
There is no consumer market available yet for sub 0.25uM devices
so there is no real need to produce advanced products with limited
short/mid term profit potential.
300mm pilot lines will take 1-2 years to become fully functional
and the equipment sets required will undergo numerous productivity
and enhancement upgrades. It would be a mistake to buy the newest
300mm equipment till all the bugs are worked out or the cost of
improving and upgrading it to a "production worthy" status would
be cost prohibitive, both from an equipment and engineering resource
perspective.
Will fabs upgrade? Eventually. Why will they upgrade? To improve
process capability, yields, and PROFITS. Is it imperative that new
fabs be built today to handle 300mm and sub 0.25uM device geometries?
NO! But in 1-2 years YES.
Touche',
BB
Oh, by the way, I have this insatiable craving for a baked potato.
(Spelled with an O not oe) |
