Thanks for the question. We are discussing this but given the pretty regular commentary about failures in Class II and III we aren’t seeing high demand amongst users. The work is fairly trivial from our end but the demand isn’t high right now so we parked it for a bit while we focus on things with a broader impact. For example, we are aware that Class III is not yet accepted in virtually any critical systems, nor finding widespread adoption of even Class II for the majority of applications. I’m confident this will change but even with WSP devices, I’ve yet to see many pin counts high enough to not favor staggered uVias. I would hope to see better evidence outside of direct wire bonding and the like that led IPC to update the 2226 guidance and EIA or others (perhaps UL) pick this up with mfgs to ensure higher reliability before prioritizing this above other work. I am also not aware of FAA or MIL would consider this (yet) and would expect that to be a lagging indicator but also a well-researched (one would hope) analysis of the metalization issues at the junctions.
If I can ask, what is the application and is it something where staggered uVias wouldn’t work? I am assuming it is but thus far I have seen a bit of high temp MCM (nothing packaged or overmolded rather), some other applications going directly bonded and some memory stacking, but not much else. I am genuinely curious because it could indicate a trend that leads us to raise the priority and do it right. I have seen some other implementations but many are pretty weak or downright misleading. If the aim is to find THE tool that will do it right, then Id suggest we chat and you connect with my team because we are always looking for the real-world use cases to ensure we can build things a wider range of people might then start adopting with confidence.
thanks,
matt - autodesk