Jay,
Your post 7372 was a crucial one.
I read it about 35 minutes before the c.c.
I was able to log off, dress, jump into my car, drive to work, not get a speeding ticket, rush into my office, log onto the call and take notes. I was hoping others had done the same because, frankly, I am getting tired of hearing myself talk on this doggone thread.
These are the notes I took...
The press release describes a "Memorandum of Understanding" or a contemplated agreement between SanDisk and Toshiba. It is not finalized and could fall through. It is hoped that the details will get worked out and a final agreement signed by January of 2000.
PLEASE READ THE PRESS RELEASE AND THE RECENT QUARTERLY REPORT OUTLINING THE RISKS ASSOCIATED WITH THIS INVESTMENT.
Specific risks may include, but are not limited to: 1) *the potential lack of competitiveness created by the cooperative effort (complacency?), 2) *failure of the products to be cost effective or reliable, and 3) *increased near-term start-up costs due to increased R&D/product/control/market development costs in advance of product manufacturing and release.
The joint venture is quite positive for SanDisk as it incorporates the possibility of capturing and holding a significant portion of the market share and allows capacity to increase substantially over the current allowances and limitations of the fabless model. This is because the flash memory market will grow from about $600 million to the range of $10 billion in the next 5 to 6 years. SanDisk's current fab arrangements in Taiwan alone could not meet this demand.
SanDisk hopes to control production capacity by continuing with the current NOR/256 Mbit/D2 technology at UMC and adding the application of Toshiba's NAND architecture and advanced lithography skills via this joint venture.
The joint venture was described by Eli in 3 separate compartments:
PRODUCT DEVELOPMENT
The joint venture between Toshiba and SanDisk will produce 512 Mbit and 1 gigabit flash chips initially, with plans for 2 or 4 gigabit products in the future. This be accomplished by fusing Toshiba's NAND architecture and 0.16, 0.13 and 0.10 micron processing techniques with SanDisk's patented flash technologies including D2/MLC and advances in controller functionality. The target of these advance wafers will not only be the SecureDigital card, rather it will be employed across the spectrum of the SanDisk product line.
PRODUCTION
The production of the advanced flash wafers will be split 50/50. Currently Toshiba has 0.16 micron capability in the Yokkaichi fab, however much of the production is already spoken for. I believe any incremental increase in production or perhaps production gained from the phasing out of old production contracts may account for the intial production of the joint venture. This should happen as early as 2000. Production will likely be constrained ("relatively little" output for the joint venture) initially because current production at the Yokkaichi fab has been geared to current demand. SanDisk would pay a given price for wafers coming out of the Yokkaichi fab, because it has no capital investment there.
The second area of production will be in the upcoming Dominion fab in Virginia. A clean room will be leased in the second module (Currently only 2 modules exist. The first module is run mostly by Toshiba (128Mbit DRAM) and partly by IBM.) SanDisk and Toshiba will make significant capital investments for the lithography equipment and other processing tools. The magnitude of this investment is $150 million each during the first 18 months (beginning January 2000) and upto $400 million each to bring the second module to full production of 20,000 eight inch wafers per month.
Thereafter, the Toshiba/SanDisk entity can jointly invest in further fabs as required to meet total market demand. In particular, the Dominion site has room for further expansion.
As stated earlier, in 2000 and 2001 there will be an increase in R&D spending. By 2001 we should start seeing a more significant portion of SanDisk's flash memory being of the NAND variety.
MARKETING
Each company will market its own products and determine how the allocated lots of wafer will be utilized.
This is what I recall to be true. I may not have all the facts down pat. I was writing as fast as I could.
Ausdauer |
