You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
I figured that I really have poor judgment (or maybe I'm an idiot, I don't know), so want to play safe and ask. I guess your take is that they are not going to have it then.
In an ideal world, Intel would have been permitted to sell/transfer the foundry assets to SMIC and that way, you would have credible competition with TSMC. This would mean that the US government would lift sanctions on China wrt to EUV.
When folks compare 18A to N2, one has to keep the following in mind:
1. 18A is targeted for HPC per Intel. That filters out the mobile customers - a good chunk of the market.
2. HPC customers will usually have larger die compared to mobile customers; backside power adds to complexity.
3. Intel is new to this foundry business; there is a steep learning curve involved.
Fabless companies will be loathe to take so much risk at once. It won't be surprising if they're just waiting to see how Panther Lake pans out before jumping in. It is, after all, a high-volume product.
There is no denying that if Intel pulls off Panther Lake in 2H'25, it will go a long way in restoring its technology leadership credentials, from the days of 10nm ruin.
I, for one, am cheering both Intel and TSMC; competition keeps life interesting.