Very interesting times. It's really hard to comprehend how 20 years ago transistors were ~35x bigger. I personally think I'll be avoiding sub-7nm products for a little while, until I hear more about reliability and data integrity. Engineers might be able to work around the limitations of physics, but you can't cheat or change physics.

schmidtbag:

Very interesting times. It's really hard to comprehend how 20 years ago transistors were ~35x bigger. I personally think I'll be avoiding sub-7nm products for a little while, until I hear more about reliability and data integrity. Engineers might be able to work around the limitations of physics, but you can't cheat or change physics.

They'll never put it into production unless it has certain level of reliability. As recall of millions of chips from very expensive devices may put foundry under. Till now we have not seen any significant chip degradation over time. Biggest problem is to get defect-less chip. And they seem to be overcoming it.

EDIT: Nevermind

schmidtbag:

I understand that, but it's not like they have had years to test this. Normally I don't care when die shrinks happen - that's a pretty regular thing. The part that concerns me is how some have speculated 6-7nm is basically the smallest silicon-based electric transistors can get before stuff like quantum tunneling becomes a problem. So, I'm a little skeptical.

Yes, but if they go for mass production, they trust it. Maybe High clock OC may become issue. But on other hand, those devices will be 1st in lower clock, high efficiency phones, tablets 😀

Fox2232:

Yes, but if they go for mass production, they trust it. Maybe High clock OC may become issue. But on other hand, those devices will be 1st in lower clock, high efficiency phones, tablets 😀

I just realized that these don't appear to be traditional silicon-based transistors, and were in development for years along side traditional transistors. So - I'm pretty sure I'm wrong about what I was thinking anyway.