Intel $6bn acquisition of Mellanox !!, page-189

4DS is about data retention
WBT is about manipulating the data to serve application spinning off IOT, 5G and machine learning.

That's what is on the brochure. And its all a load of guff. The use cases for reram are very broad and will map across a number of domains. There is no reason why a "working" 4DS chip couldnt do a perfectly adequate job where WBT is "targeting", and vice versa.That WBT are playing around with AI and in-memory processing is not the main aim, its still about writing, storing and retrieving information efficiently. Same for 4DS.

I use the term "working" for a reason.
It's all theoretical right now. There are many unanswered questions from both camps despite the enthusiasm on HC forums at each announcement.

They all really come down to one question though:

"Can either of them get performant, high yield, commercially profitable chips from their processes."

Nothing else matters in the end, no matter how much sparring on relative theoretical merits or talk of target markets is done here or on 4DS. For instance, you can have a million year retention life but if you are only getting 5 chips off a wafer then you probably arnt going to be around for very long. And fabbing on commercial equipment is great, unfortunately anyone fabbing is doing it on commercial equipment, not just 4DS. And doing it on that in no way implies they are getting commercial yield. I may have a ferrari but only drive it at 60km/hr. This is like on the test track right now, adjusting parameters after each test lap to see what works best for the actual race.

And none of the key information to answer that question above in bold have been clearly delivered by either party with the exception that we know WBT has some sort of packaged silicon out for test with some groups. That does not mean yield is high, quality is consistent, they are complete chips or potential costs are reasonable. We know 4DS have working cells, we dont know what that means really in terms of being able to get a profitable product together in any way soon. Its better than no working cells, but you want to hear then getting the saw out at some point and packaging rather than bench tests.

Sorry, I see a lot of very naive exhuberance here. Both makers are progressing well but I agree its a long road to commercial product.