The Yangtze River Riddle

It’s often said that if you want to get into a technology as a newcomer you have to aim-off so that when your product hits the market, it is technically ahead of everyone else’s.

However China’s flagship chip company Yangtze River Storage Technology,  tasked with getting China into the memory business, is looking to have a 32-layer 3D NAND chip in volume production in 2019.

Samsung and Toshiba say they started production of 64-layer 3D NAND last year.

To have, as your target, the aim to be a product generation plus three years behind the state-of-the-art when you enter a new market seems a trifle unambitious.

When the Japanese made their big DRAM push in the 1970s they developed the 64k, 256k and 1Mbit generations simultaneously instead of following the usual industry practice of using the profits from one generation to pay for the next.

When the Koreans entered the game, Samsung simultaneously built three fabs for three successive DRAM generations.

Yangtze River’s CEO, Simon Chang, says his aim is to break the market stranglehold of Samsung, Hynix, Toshiba and Micron in 3D NAND.

Obviously he’s not going to do that by selling obsolescent memories, but Chang is also saying that he expects to catch up with the industry leaders in the technology by 2020 – a year after he makes his first 32-layer NAND chips.

Assuming the leaders are on 128-layer chips by 2020, Yangtze River has to master two 3D NAND generations in the next three years.

Without any obvious source of the technology open to them, that’s a massively tough call.

So either Yangtze River is not telling the whole truth about its strategy, or it is heading for a fall, or it has been guaranteed a local market for trailing edge products.



  1. You’re right Marijn, it does make some sense but, compared to how the Koreans and Japanese did it, it’s lesss science-based, it’s less rigorous and it’s less ambitious.

  2. An excellent point Duncan, that does help explain a lot

  3. “or it has been guaranteed a local market for trailing edge products.”

    Given Apple’s aversion to putting a decent amount of Flash in anything but their top of the range products they shouldn’t have any difficulty meeting most of Apple’s requirements 🙂

    More seriously there must be a huge market for Flash in £50 Smartphones, Tablets etc, none of which are cutting edge applications.

  4. “or it has been guaranteed a local market for trailing edge products.”
    That’s what happened, same for Chinese foundries SMIC and HuaLi. They explicitly mentioned they do not have to be competitive on cost, they only need to deliver a working process. The party will ensure it sells.

    From a government perspective this makes sense to me, it’s a temporary economic cost to get independent of other nations that have stated they are not afraid to use their power.

Leave a Reply

Your email address will not be published. Required fields are marked *