> But Intel made it work, produced this stuff, put it on the market… and not enough people were interested, and now it is giving up, too. That sounds like a du

> But Intel made it work, produced this stuff, put it on the market… and not enough people were interested, and now it is giving up, too. That sounds like a dubious claim to me, at least if we take the hyping up at face value. If Optane was really that awesome, why not build a database on top that's either much faster or much cheaper than what we have now? Then sell that to a cloud vendor. The end user is only interested in the wire protocol (and up the stack from there) that they speak to the database, and not the underlying OS. I cannot imagine Intel going to one of the three major cloud vendors, offering a 10x improvement (in either cost or performance) for one of their larger offerings, and the cloud vendor simply saying "nope, not interested". Maybe the practical improvements simply weren't there?