Ann: Weebit signs first commercial deal with US-based SkyWater, page-135

I think the problem is the huge amount of logic being made relative to the fab capacity to make it. Fab line times for production runs are the problem. Good thing here is at 130nm, they may be in the IOT sweet spot according to WBT but I dont think they'd be in the prime demand for fab lines, which is probably in the 40nm to 28nm territory. This may be good thing coming off a relatively unsophisticated line with a more sophisticated memory. Demand for fab time may be much lower than the spot where the fab bottlenecks live.

If you are talking the logic shortage perhaps a little, in that an easier to produce memory that integrates at the right density in less passes and requires less masks will take less time to produce, thus freeing up time on the production line. Will it be that big a contribution, probably not but it might help.

However, there have been a number of recalls of tech due to the way the current flash based control boards (and there are a number across an EV) handle writes and spread the write load. They simply wear out too quick when they are doing complex stuff. A better reram with much longer duration might stop this problem dead in its tracks.