Share
911 Posts.
lightbulb Created with Sketch. 802
clock Created with Sketch. 12/02/19
00:37
Share
Originally posted by You Beaut:
↑
I think that you understate:the distinction between filamentary and non-filamentary reram . All the filamentary reram companies (such as wbt) have ended up competing in the embedded memory space ("convenience stores") due to physical limitations associated with having a conductive filament. 4ds is going after the much larger market, and effectively virgin territory (if you discount 3D Xpoint), of storage class memory ("banks").the value of a commercial partner - provides independent validation as well as insight (eg what has previously been trialled/failed) and guidance as to what the market really wants. WBT may well have given away 40nm 1MB arrays to third parties, but responses to date have been from academic institutions, or from service providers that wbt will need to pay to develop its product, rather than serious industry players interested in using/licensing/buying its technology. This tends to suggest a collective yawn from the industry as they are either aware of the limitations of, and limited market for, filamentary reram or have their own filamentary reram technology (including WD) - or just simply have doubts about wbt's tech.
Expand
Okay so that's my biggest question, and you've almost answered it. If they produce a megabit test chip by the end of 2019 and it works. Where to from there? License the tech? Like, is that game over for the speculation then? At that point does it become the next HUGE leap in memory computing and people will be chucking billions at us? Or is there still lots more to do after that?
