Exiting commodity DRAM to concentrate on NAND flash (2001)

Choosing to fold up a world-leading business

The crux of this decision lay less in the position of being number one in the world than in reading clearly the character of the market that position sat in. Commodity DRAM demanded enormous investment every generation to keep shrinking the circuitry, yet the more the product approached a standardized good, the harder it became to hold the initiative on price. As the era in which Toshiba led at the one-megabit generation gave way to one in which Korean and Taiwanese makers closed in through mass production, there was a certain logic in letting go of a commodity that no longer earned and shifting resources to the NAND flash memory it had invented itself. Deliberately folding up a business built on past glory is a decision that companies weighed down by their own success find hardest to make.

That said, the same pattern of “selection and concentration” would show Toshiba another face in the years that followed. The instinct to funnel resources into the businesses that earned was carried over into the vast investment in nuclear power, which eventually brought enormous losses. Cornered, Toshiba would let go even of the NAND flash memory business it had kept back when it exited DRAM. The judgement of what to keep and what to discard forever carries, back to back, both the power to read a market’s character and the danger of concentrating the company’s fate on that bet. The withdrawal from commodity DRAM stands as an early step toward a Toshiba that chose its businesses, and it leaves open the question that its later concentration and divestment would pose again.