Towards 2 nm chips

Towards 2 nm chips Modified microwave oven cooks up next-gen semiconductors uses controlled position of the standing wawes...

formatting link

Reply to
Jan Panteltje
Loading thread data ...

IC scaling is leveling off. Costs are no longer declining with scaling, and the design and mask set for a high-end chip can be a billion dollars. Not many chips justify that.

Maybe we don't need denser memory or faster logic.

formatting link

Reply to
jlarkin

On a sunny day (Sat, 10 Sep 2022 10:09:45 -0700) it happened snipped-for-privacy@highlandsniptechnology.com wrote in snipped-for-privacy@4ax.com:

Sure Moore's law does end by definition at the molecular / atomic scale. And indeed low complexity chips are - and will be needed .

What's not mentioned is the wars that will come.

We now wonder: 'How did those Egyptians ever build those Pyramids?". Maybe many years after the total breakdown of current society due to nuclear wars, climate change, geological effects like volcanos, maybe thousands of years from now, "How did they ever go to the moon?" "How could they possibly have made those cellphones?" (plenty of those to dig up by archaeologists).

We, the human race, seem to be at some peak close to self-destruction. US war machine sales force creating tensions everywhere to make wars and make wars last longer. Mass migration of lower IQ populations flooding developed countries..

So, like we now look back at the pyramids....

Back to 'tronincs.. many things done with advanced chips can be done with simpler components. They mention Microchip, I can do a LOT with a few PICs, a few transistors... Raspberies have replaced PCs here... powerful enough but no, I am not processing bitcoins and not gaming.. But even raspberries are unobtainium now. Still wonder how I got that Pi4 8 GB.. must have been the last one!!

Reply to
Jan Panteltje

That is a thought I have had. I know we document tremendously the science and the technology of making chips. But sometimes I wonder if we also have an oral tradition that is not well documented such that a thousand years from now people will look at some of our inventions and wonder how we could pull that off.

It seems pretty clear that with the semiconductor shortage lasting so long, we must be winding up the integrator on the control loop and we are going to see massive oversupply and excess production capacity before too long. I figure it will be going sideways as soon as the first fab conceived during COVID starts shipping chips. Every plant after that will be fuel on the fire.

I don't recall just when it happened, but the memory sector had a similar shortage, which resulted in reorganization where companies sold off their memory groups to form super-memory companies. Then the glut came and they were all going bust from excess capacity. The original companies still owned shares of the new companies. The end result was splitting up and selling off of some of the new memory companies.

One part of the semiconductor business is consistent. Every problem can be solved by buying and selling business components more so than electronic components.

Reply to
Ricky

Of course we do. We may be running out of the mass market applications that can pay for them.

That does emphasise that the cost per transistor has levelled off, but this misses the point that a faster (which usually means smaller) transistor can do more work, just because it is faster, so the cost per mathematical operation hasn't levelled off yet.

Reply to
Anthony William Sloman

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.