Very interesting! Hopefully this comes in time and yields the expected results!

I don't get it - why haven't they done this like... 20 years ago? Also, it's a pretty bad sign about their power consumption if they're trying to take a whole new (and likely costly) approach on a tried and true method that they've already attempted to meddle with in previous models (9th gen if I remember correctly is when they started to shave it a bit thinner).

schmidtbag:

I don't get it - why haven't they done this like... 20 years ago?

the sand purity required isn't a laugh.

When eyeing 7 and 8 GHz, now that 6 GHz was reached, Intel needs something that can withstand even more power and heat.

Astyanax:

the sand purity required isn't a laugh.

When it comes to the doped silicon, absolutely, but I get the impression this is an add-on substrate to replace the resin they've traditionally used. There's no way Intel would waste that more-valuable-than-gold pure silicon on a thicker die for ever-so-slightly better cooling performance; I'm sure 99.0% pure silicon dioxide would be plenty good enough for this application.

schmidtbag:

I don't get it - why haven't they done this like... 20 years ago? Also, it's a pretty bad sign about their power consumption if they're trying to take a whole new (and likely costly) approach on a tried and true method that they've already attempted to meddle with in previous models (9th gen if I remember correctly is when they started to shave it a bit thinner).

everybody used to use ceramic packages, I'd imagine they moved away from that purely for cost reasons.