Toshiba’s semiconductor saga has a new twist. The Japanese conglomerate is reportedly splitting itself into three companies focusing on infrastructure, devices, and semiconductors. According to a Nikkei report, Toshiba is on corporate drawing board, again, this time to set things up once for all.
The company’s infrastructure business will handle nuclear and thermal power generation while its device unit will manage the disk drive business. That leaves the semiconductor unit with flash memory and discrete products.
Figure 1 One of Toshiba’s claims to fame in the semiconductor industry is the invention of NAND flash memory in 1987. Source: Toshiba
At this new crossroads for Toshiba’s semiconductor operations, it’s important to understand what has been going on here during the past few years. First, in 2017, Toshiba sold off more than half of its stake in the flash memory business to a Bain Capital-led consortium to pay for restructuring. Next, it spun off the remaining 40% of the memory business into a subsidiary named Kioxia. Pronounced as kee-ox-ee-uh, Kioxia combines the Japanese word kioku meaning “memory” and the Greek word axia meaning “value.”
Here, it’s worth mentioning that Western Digital, Kioxia’s production partner and a joint investor in Toshiba’s memory chip plant, has proposed a merger that’s not likely to mature. However, it has pushed back Kioxia’s initial public offering (IPO) plans. Now, if Toshiba splits its semiconductor business into a separate entity, it will most likely be the holding company for Kioxia.
Figure 2 Toshiba Memory Holdings Corp. was officially renamed Kioxia on October 1, 2019. Source: Toshiba
Then, in 2020, Toshiba closed its LSI business that developed products like image recognition sensors. The move was followed by Toshiba’s sale of two semiconductor manufacturing plants to Taiwanese fab UMC. Toshiba has been selling semiconductor business assets to help offset massive losses incurred by its U.S. nuclear power business.
After all this wheeling and dealing, it’s a moment of truth for the Japanese tech giant, and as a result, a new chip company will soon be on the semiconductor block. However, the above labyrinth of news shows that merely spinning off the semiconductor business won’t be sufficient. It’ll only be the first step; there will be several challenges waiting for the new company.
First and foremost, how will the synergy work between Kioxia and the rest of Toshiba’s semiconductor business? Will the new company decide to go fableless or fablite? Will it have sufficient capital to reinvigorate its engineering operations and potentially make a few acquisitions? Unlike Hitachi, Mitsubishi and NEC, Toshiba has been late to the chip spin-off party, which brings it a sufficient learning curve to avoid mistakes as a newly independent semiconductor outfit.
In retrospect, Toshiba should have done it long ago, but as they say, it’s never too late. Focus, speed, and leadership will be critical for the new semiconductor company whose name is still to emerge.