Ann: 4DS Australian Open Briefings Presentation, page-328

At the risk of re-igniting the retention debate yet again, the quote from Peter in his Sydney presentation when discussing tune-ability on slide 6 (Can trade off longretention for reducedendurance) was "We don't talk about it but we can emulate non-volatile memory."

We can postulate on why they don't talk about it (e.g. Commercial, confidentiality, lacking certainty or theoretically possible but not productizable) and exactly what "emulate" means in this context. But, IMO, Vinneh et al have been correct all along. We aren't being told the full story on retention. Having said that, I'm guessing there are good reasons why they are being economical with this information and since I'm not in possession of all the facts I can live with their strategy.