What would be really interesting to know is the preference order in their list of emerging non-volatile (NV) memory types that they plan to investigate. Will that list be topped by ReRAMs, Magnetic MRAMs, Feroelectric FeRAMS? Of ReRAMs will they opt for mechanisms based on Bulk, Filament, electrochemical bridge, electromigration, vacancy movement or will they take their chances with phase change (PCM). As well as the management team mentioned by one poster here they will need a massive resarch effort and deep pockets to come out on top. They will also need to carefully study the history books that chart past NV memory development efforts (Outside of Flash) and be able to separate the snake oil salesmen from reality.
I am not sure what manufacturing capabilities IMETU has but GigaDevices is fabless and therefore has relations with foundries (possibly SMIC?) and could probably make arrangements to run R&D wafers there at reasonable but not leading-edge geometries.