TSMC on Track to Begin Volume Production of 3nm Chips in 2022

[the8thark]

picometers
then
femtometers
next
attometers
and even smaller
zeptometers
yet even smaller
yoctometers

I think the main question is how does physics work with something smaller then the invidivual atom and compounds?
Do you need a rethink on fundamentally how everything is done once you hit atom level sizes?

 

[cmaier]

I think the main question is how does physics work with something smaller then the invidivual atom and compounds?
Do you need a rethink on fundamentally how everything is done once you hit atom level sizes?

No. Just because it‘s a 3nm process doesn’t mean that anything is 3nm in size.

At some point transistor topologies will have to change again to enable better gate control (like they did a few years back when they went from flat FETs to FINFETs), and at some point further down the road from that they will have to get more creative. But there’s quite a ways to go. You can even keep the size the same and use bandgap engineering to double (or more) the performance - all it takes is the incentive to do it, which will happen some day when they can’t just make things smaller.

 

[SFjohn]

What comes after 1nm?

Angstrom (Å) / picometer

 

[crazysword]

Already mentioned ijn another thread, but hey, again:


not gonna protect Intel for „slouching“ around vs the competition, but could please EVERYONE here going like : „duh Intel is so behind - they are still on 14nm++++++ - they must be the dum.... company on the planet“ please back off.

7 vs 10 vs 14 nm has NOTHING to do wit Transistor size (anymore). It is simply just a remnant of earlier times and the „nm“ term is only used to name generational jumps/improvements. Whether it is a declining number (14-10-7-5-3....) or gen 1,2,3,4,5, or just simply adding (those friggin ) „++++“ -it simply doesn‘t matter anymore.

For clarification: Der 8auer actually took Intel and AMD under „a“ microscope - and guess what: Size and evolution DOES not really differ - so please bash Intel otherwise/correctly.

Video part 1:

Video part 2:

Video part 3:

Enjoy learning about „14“ vs „7“ nm

This video is not proof that the different foundry processes "does not really differ".

We have objective statistics of millions of transistors / mm^2. TSMC 5/7nm are significantly ahead of Intel. Not only that, we can estimate power efficiency and Intel is also significantly behind TSMC.

All Der 8auer showed is that some of the features are similar in dimensions.. that's it. It does not change the fact there are significantly higher transistors density and better performance with TSMC.

 

[the8thark]

No. Just because it‘s a 3nm process doesn’t mean that anything is 3nm in size.

At some point transistor topologies will have to change again to enable better gate control (like they did a few years back when they went from flat FETs to FINFETs), and at some point further down the road from that they will have to get more creative. But there’s quite a ways to go. You can even keep the size the same and use bandgap engineering to double (or more) the performance - all it takes is the incentive to do it, which will happen some day when they can’t just make things smaller.

What you said is very true but it doesn't invalidate hat I said. I agree that the process size and the end result sizes often don't match up. Still eventually the whole size of an atom might be a physical limit they all need a work around for.

 

[cmaier]

What you said is very true but it doesn't invalidate hat I said. I agree that the process size and the end result sizes often don't match up. Still eventually the whole size of an atom might be a physical limit they all need a work around for.

They’ve been saying that since at least 1996. But the people who say that are not people who actually design chips, for the most part.

 

[chucker23n1]

That would be nice if true, but Qualcomm bought Nuvia headed by the former Apple Silicon designer. Their chips boast the same performance as Apple’s M1’s naturally.

They don't have a chip. They might have some interesting prototypes, at best.

Apple is in no position to be cocky as the competition is only waking up, and the market for non Apple computers is far greater.
Competition is already nipping at Apple’s heels.

I don't think Apple is snoozing in this regard.