My expectation for the Chinese DRAM manufacturers is that they will optimize at first for the commodity DRAM market, focusing for the slower grades of DDR-4 in smaller DRAM sizes. I'd expect 4 and 8GB sticks from them at first that are modestly specced (think moderate to high CL 2400 and below). Given the volume in that section of the market, even modest markups from Samsung and Hynix will be under a lot of pressure quite quickly.
Samsung and Hynix know this quite well. They aren't stupid. They are aggressively building out capacity on the high end stuff HBM2, DDR6, high spec DDR4 in an attempt to stay relevant and profitable against very low cost products from China that are expected on the market soon. The profit taking on the top end right now is funding that expansion. Given how far behind Samsung Hynix is right now, I fully expect them to financially fail under the brunt of the Chinese products in the market and to eventually be purchased by those same Chinese firms.
the next few years in the industry are actually quite clear. Samsung will be blazing a trail at the top end of the market in a fight to keep that segment of their business alive. Hynix will fall. The Chinese firms will gobble up the makers that fail to get additional technology and IP for western sales and continue to make volume in the low and mid markets. We NEED Samsung to survive to keep the Chinese companies from gaining a monopoly on the market themselves. We NEED the Chinese firms to put pressure on Samsung both by volume and technology development to keep prices in the middle of the market manageable.
The end game is unfortunately already written. Eventually, Samsung will hit an expensive technology roadblock, somewhere around 7-3nm, and will go broke trying to get past it. The chinese firms will eventually catch up to them, break them financially, and buy what's left once it gets spun off. This won't just happen in the DRAM market either. They are making a lot of headway in the CPU/APU market with the VIA/Centaur team. Yes, it isn't competitive TODAY with even mainstream Intel/AMd products, but, in a few years, it will be. In 5 years time, they will have most of Asia covered in inexpensive, competitive, mid-market products that will guy a lot of Intel and AMD's volume product market. Without that volume, they will have a hard time funding the rest of their development projects. AMD may be in a better position without the pressure to keep their foundries in business. But, in a decade, I don't see anyone still in the market except for Chinese companies.