What a lot of pundits / punters don't realise is that CEA Leti is a cutting edge / global leading foundry ( now partnered with Imec just out of interest as well as many, many other globally leading companies in the field of semiconductors) with 200mm / 300mm wafer production capacity upon which all wbts' chip / industry verification processes have been completed ......validation / verification completed to industry best practice. They (LETI) have just assisted ST Micro into 28nm FDSOI production ...huge!!!!! Further to that - in the knowledge that wbt intend initial production on a 200mm wafer - all development work and test validation processes have been conducted on a 300mm wafer /fab equipment to facilitate an expedient move to 300m 28nm gate chip production. ( as per last release to market)
XTX independently validated the chip in record time and verified to industry benchmark ( not test lab) standard that this thing is exceptional - ffs without that verification by them ( thankyou Jackson Lam ) there would be no acknowledgement in industry circles that wbt's stuff had even left the test bench.
Perhaps no bells . no whistles - but imo once that first deal drops it will come with a rush.....investors will really start to understand just what these guys have put together and the capitalisation will be realised exponentially and accordingly with the cheapest cost per bit production in the industry (wbt) - a big claim but a fact nonetheless. Thats what silicon is - ubiquitous and the most readily available material on the planet.
The other product streams will ramp up into commercialisation the moment the fab deal is announced ( which is very close as we know imo) ...hmmm, March / April 2021.....sensible comment suggested in December 2020 that 3 to 4 months into 2021 something of substance would be announced...its so close now - patience is all that is needed for this thing to go exponential. That's what the announcement about the guys getting on with the 28nm gate is all about - were not sitting here twiddling our thumbs - the work continues into the scaled nodes ( Its a good thing to keep to the back of your mind that 5nm gate was test verified by James Tour in 2015 ffs - 5nm !!!!! no wonder the industry couldn't quite believe what the good professor was saying.
Most informed investors here know all the product under development by the weebit team - well this randomness thing that Leti have stumbled across is groundbreaking - another disruptive breakthrough on a level that most cannot even conceive of.....
Why do I say that - well coming off the test bench in system verification you need to make sure you get to market as quickly as possible, you do so with a quailty product that has been tested ( and the process flow around testing is enormous) and your an affordable option. Miss any one of those and you sink ( risk mitigation) .....well simply verication of design ( the product has been tested as above) is 30 to 35 % of cost despite all the automation that is within the industry ....verification is the bottle neck if you will - this MCMC randomness leveraging in Reram potentially reduces that cost an opens edge learning access to levels not seen in the industry, ever. That is to say - it does away with process written test algorithms for verification - (2000 seem to be the limit) - it has gone beyond the ability and limits of algebra written paths of verification. - so leveraging randomness potentially streamlines the verification process in ways unimagined previously - and when you consider the verification process happens multiple times across the development path it signals faster path to market with a higher quality product capable of higher write/ store and production functions. Look it up - it may may well rewrite the definition of disruptive in AI .....and thats a big call - but the data and evidence support the claim.
Some others claim Reram production in competition - sure, bring it on - the reram / product developed here is beyond what the competition has developed - I stress this is my opinion - but again the evidence on the bench and verified by industry (Leti - a leading global foundry -and XTX independently) evidence this fact....easy to loose sight of sure, but no less ready.
The chip these guys have is ready - its now striking the right licencing agreement - perhaps some could look to the deal crossbar struck with SMIC as a comparative ...or Rambus' deal with Fujitsu back in 2012 ...there are good precedents on what is possible in the space.
Looking forward to the news flow as always views here are my own and open to scrutiny and due diligence of all who may be interested.
(20min delay)