In terms of yield, volume, and delivery. TSMC will be shipping 2nm in late 2025, while TSMC US will be doing 4nm, a variation of their 5nm in the same time frame.
That is 3-4 years behind. I guess TSMC could make TSMC US roughly 2 years behind Taiwan. And there are plenty of usage for these nodes. I mean most GPU sold are still in 4 / 5nm.
For US to be back in the lead in terms of node, they could only count on Intel.
After reading this, I was interested to see how Intel is doing with their 2nm process. It turns out that they've shelved 2nm and moving ahead with 1.8nm (18 Angstroms, hence the "18A" name). This would be really great for Intel and the US.
Does this mean we can soon approach a point where the leading chip designs won’t be sent abroad for fabbing? Would be a national security miracle
Define "soon" and "leading".
In terms of yield, volume, and delivery. TSMC will be shipping 2nm in late 2025, while TSMC US will be doing 4nm, a variation of their 5nm in the same time frame.
That is 3-4 years behind. I guess TSMC could make TSMC US roughly 2 years behind Taiwan. And there are plenty of usage for these nodes. I mean most GPU sold are still in 4 / 5nm.
For US to be back in the lead in terms of node, they could only count on Intel.
After reading this, I was interested to see how Intel is doing with their 2nm process. It turns out that they've shelved 2nm and moving ahead with 1.8nm (18 Angstroms, hence the "18A" name). This would be really great for Intel and the US.
https://www.tomshardware.com/pc-components/cpus/intel-18a-pa...
archive/paywall: https://archive.is/MckxG