Chip War: The Fight for the Most Critical Technology [video] (opens in new tab)

I think all the commenters focus way too much on litography and chasing nanometers. Yes they're an important segment of the market, but they're just a small part of the market. Even 80nm-scale litography is increadibly important, volume-wise, because a lot of industrial applications are on older nodes.

Furthermore there are a whole swath of types of chips for which smaller litography isn't even applicable. China is developing GaN chips quickly, and these are power chips that will waste less power and become less hot.

And apparently building a 5G modem and a sattelite receiver chip aren't trivial things either. Apple failed to build a viable 5G modem for the latest iPhone while Huawei did it, showing that analog chip design (way more complex than CPU design) should not be underestimated. Apparently BW filters (a component used in 5G modems) aren't trivial either, with very few suppliers in the world. Existing sattelite phones has clunky antennas but the one by Huawei can't even be seen. And it looks like Huawei innovated heavily in heat dissipation technology, allowing its latest smartphone to perform competitively with other flagship phones despite the chip generating more heat due to bigger process node.

The list goes on. Just focusing on nanometers and litography is really a shallow view of what's going on.

Now the talking point is 'well sure, they have 7nm. But if we cut off their duv stepper access they won't be able to maintain the steppers they have. And definitely won't be able to make euv steppers'.

Meanwhile in reality they already have at least 28m DUV steppers indigenously built and are rumored to have a few 7nm capable DUV steppers doing shuttle runs. And are doing really neat things in the EUV space taking a different direction than ASML, essentially using a particle accelerator rather than a tin laser.

It's been clear now for over a decade that every time we cut them off from a technology they produce indigenous versions. We cut them off of chips to go in a supercomputer and within five years they had the world's fastest supercomputer on indigenous designed chips with Tianhe-2.

40nm prediction must be based on that apart from EUV even DUV export will also be banned. That will be bad and China must be stockpiling everything it can before that happens.

> I think in a few years the tech war will be obviously lost.

I think the idea is to extend the no. of years lead US+ has. China will have to replicate one of the most complex supply chain ever seen. They will use all measures possible so would be interesting to watch.