@Prices - losses
I think the current NAND environment is "good" for us in the
long term. Ok, in the long run we are all dead, but you know
what I mean.
We are now seeing heavy Capex reductions at nearly all DRAM
guys and I assume, these reductions will even widen, when we
enter the beginning of 2009. When you ask me, such moves will
be seen at NAND in the next time. All these guys which hoped,
that NAND Mbit demand will increase 200-300% each year have
to look for other new growth stories, which don't appear mostly
over night. SSD is clearly one theme, but I have questioned
a fast switch months ago, not only on this board. SSDs aren't
fast enough, too small and way to expansive. The problem is,
that even with existing prices NAND guys loose money. To
drive real growth into this sector I would guess, the drives
have to fall in price by 50% or more. Lets take a look at
the harddisk market. I have looked at 1TB drives which went
to 25-30% of the price 9-12 months ago. You could buy 750GB
for a very small budget, why pay hundrets of $$$ for 32 or
64GB SSDs? I'm really sure, that there are applications where
SSDs will make a huge difference, but in the end, the consumer
is driving the REAL growth, but not at these prices - thats
for sure.
If Capex reductions will appear - hard reductions, the fast
switches and even more Mbit capacity will not appear. I'm
assuming, that newer nodes will not be used widely on huge
% of capacity, because switching will cost big money. We have
spoken about that already months ago, that the speed of
introduction of newer process nodes will slow in NAND because
of physical limitations. Now I have to add to that, that the
speed will slow once more, because huge Capex reducstion will
appear or are already in place. So, SPSN also doesn't have
huge amounts of money, sadly, but the existing gap to NAND will
get smaller and smaller once we ramp SP1 on 300mm with 65nm
and hopefully 45nm in the not so disctant future. The most
crucial topic for SPSN and for me is the introduction of Quad,
which should drive costs significantly down. We haven't heard
that much about that and I'm really hoping, that SPSN is
able to manufacture these designs on 300mm and first 65nm in
masses, which should be near on par with current 4x nm
NAND designs. Once we are in this position, the whole situ-
ation will change, also when we don't play in the same
markets. Numonyx is another piece, about which we could talk
in length.
BUGGI |
