Based on everything I have read coming from BFL over the past months, I believe they received chips in late October or early November (shortly before they posted the pictures with the chips attached to them), and they didn't work properly (or performed so sub-par as to require a revision). They're doing the clock buffer revision (and who knows what else they are revising and not admitting to), and now waiting on the fab for the new chips.
Based on what they've posted, they've
never received a batch of ASIC chips, bad or otherwise. Josh is the only one who implied that they had and Nasser quickly corrected Josh's statement.
Producing custom chips requires foundry scheduling and forces us into the queue with other chip developers (from all industries). The design is complete. Its just a matter of waiting for production and delivery.
https://forums.butterflylabs.com/content/127-bfl-asic-delays-depth-expanation.htmlThe design is complete. Period. That is where things are at. They do not yet know when the production run of their chips will happen, only that BFL should be able to take delivery of them in January (after which other things still need to happen before they can be assembled into mining rigs). Nasser and Dave's statements imply that Josh was not fully aware of where the process was actually up to before and was simply assuming that the fab had fucked up in some way. As Josh's role in the production cycle doesn't begin until it's time for the chips to be sliced and packaged, it's easy to see that he mightn't have been kept up to date on every little thing which has been going on up until now and has been making statements based on assumptions he's drawn from what information was available to him rather than hard facts.