Regarding 4ds. Your testing infers to the actual materials being...

Regarding 4ds. Your testing infers to the actual materials being tested and passed to actually make it into the environment of fab, if im not mistaken.

Where as,

"“Weebit Nano’s SiOx ReRAM technology can be manufactured using existing processes with the existing equipment in production lines, as opposed to other technologies that use non-standard materials. Consequently, you can see why we believe we have significant advantages to harnessing this massive market opportunity. Using standard materials has enabled us to achieve this milestone significantly quicker than any competitive ReRAM technology”,
added Mr Hanoch.

The 1Mb initial tests, conducted in CEA/Leti facilities in Grenoble, France, demonstrated the capability of addressing and programming nearly all of the memory cells. These are very impressive results for a first batch, and are especially encouraging as they were achieved without using performance improvement techniques such as redundancy, which are commonly used in the memory industry to achieve programmability of 100 per cent of the array cells.
Weebit Nano recently announced it has extended its agreement with Leti to further develop and optimise its ReRAM memory technology. This will allow the acceleration of the optimisation process that will focus on improving the quality of the 1Mb array, bringing the technology to the level needed to move to a production level fab (where silicon components are manufactured).
The Company will continue to provide updates as further progress is achieved."

So i see optimisation.