Showing how much at risk that $5 billion is, DRAMeXchange pointed out last week that three bit per cell (TLC) NAND flash prices fell up to 8% this month, while multi-level cell (MLC) prices fell up to 3%.
Very low NAND prices during the fiscal year 2009 were the main reason for the $3 billion loss made by Toshiba's semiconductor division that year.
The effect of the fifth fab will be to increase Yokkaichi's NAND production from 260,000 300mm wafers a month to 360,000 wafers a month. The intention is to have the fab running wafers in the Spring of 2011.
Toshiba is a company with a mission. It has been trying to close the market share gap with its NAND licensee Samsung.
In Q1 this year the gap was 38.5% Samsung, 33.5% Toshiba
In Q409, the gap was 39% Samsung, 34% Toshiba.
A couple of years ago there were 12 percentage points of market share between the two companies, with Samsing on 40% and Toshiba on 28%.
In 2007, Samsung was on 45% with Toshiba on 28%.
So the gap is remoreselessly closing helped by Toshiba's success in its commercialisation of 3bit-per-cell (TLC) technology
Which is probably why Toshiba is keeping its foot down on the investment pedal.
But does it make commercial sense?
When did anything in the memory business make commercial sense?
TOMORROW MORNING: The Ten Best-Selling Cellphones