60nm megabit array - tick
clearly a 'prototype' demonstrating successful manufacturing and process and functionality
breakthrough memory technology follows a process development line:
- functional single cell device functionality delivering early performance stats
- a megabit array layout delivering large data stats
- full megabit array delivering the test chips
- foundry partner tape out with integrated BEOL with all the memory elements
it seemed possible a foundry partner may have moved early but obviously we now know differently
the right test vehicle for that IMEC foundry partner is yet to be agreed upon so we do more at 60nm and move to 20nm to test repeat manufacturability and functionality at that process node.- they have to test and demonstrate successfully to industry that it will scale
the full megabit array at 20nm then delivers the large data stats needed
If successful it will prove the technology and manufacturability presenting the foundry partner with the right 'test chip' tape out vehicle.
IMEC is the stepping stone to that 'real' fab environment and they say 'when' ....
and 'when' will mean that the 20nm functional test chip (if successful) will deliver an order of magnitude improvement delivering real disruption capability - in a 'killer' application yet to be demonstrated if and when the above has played out successfully -
The big boys ( google or others) won't move on it unless these factors are there ready to exploit and leverage into ongoing revenue through successful production run from the foundry partner - there is then something worth selling
i still wonder why a 'salesman' was required ?
its way to early - but I could be completely wrong about all this
hope everyone has a great xmas break
