I suppose this is a correction piece, concerning my previous entry concerning the future of PE in Japan by way of Bain’s buyout of Toshiba’s NAND division.
In that piece, I stated that a sub 10nm chips might be the limit, and that the jump to alternative technologies might be the single largest risk the deal has to navigate. Turns out, I might have been wrong.
According to Engadget, based on the press release by Samsung, the next batch of chips from the company will have a 10nm process, with an 8nm process well under way for mass production. Moreover, they say these chips are mere iterative improvements to their current batch of chips on the market, with the company looking to make the jump to a 7nm process soon.
What’s even crazier is that Samsung’s main rival in the industry, TSMC (Taiwan Semiconductor Manufacturing Company) is said to have the lead in this next gen, 7nm process, which means Samsung might lose a good portion of its business to its rival in the foreseeable future.
Now, the loss of business for Samsung is not something I have a good grasp on to say for certain. What I can comment on is what this sub 10nm process means for the NAND deal, and the industry as a whole. By the looks of it, the forecasts for the death of traditional NAND, even the new 3D NAND, might be premature. NAND might have some life left in it afterall.
The theoretical limits of the tech, long pegged at the 10nm mark, seems to have been broken, which means, despite the low level of performance returns on investment, established players can stay relevant for longer. Moreover, companies trying to replace the aging standard now have higher standards to meet, as a result.
In short, Bain’s gamble may have paid off just from this news alone.
Whether Toshiba’s NAND division has the capacity to compete in this new, sub 10nm world, is something that remains to be seen, but I can say with confidence that it’s not impossible. Samsung’s 8nm manufacturing process is said to be a derivative of its 10nm manufacturing process.
What I worry, now, is for companies like Crossbar, who have been researching ReRAM tech for a long time. Investors in them now have to readjust their exit timelines to reflect this new status quo. Whether it be a new funding round for the company, or even a change in exit strategy from IPO to an M&A, it remains to be seen, but there’s reason to see this as a risk in the VC space.