Source item
Open original source ↗There's 3 possibilities if it isn't BS reporting like they did with Huawei 910B just earlier this week.
1. Comparing N4 in Arizona vs N3 in Taiwan
2. Favorable mix of Apple SoC w/ fewer metal layers vs HPC in TW w/ more
3. Has newer tools like ASML 3600D vs 3400C, better overlay
To expand more. They could be comparing Arizona making a shitload of apple chips that are small and therefore have much higher yield versus complex HPC chips that are large and have more complexity in the metal layers. Those additional metal layers
You were lithography tools as well as other deposition etch etc will always have minor improvements and upgrading old tools is great but not always as good as a brand new tool. Once calibrated these new tools may lead to better yields
But also Bloomberg just fucked up earlier this week with their Huawei TSMC reporting, so I wouldn't put too much thought into this
Damn voice to text cooked my coherence on the above 3 tweets.
5 segments
Segment 1 · characters 0–279 · Original segment ↗
Segment 2 · characters 281–529 · Original segment ↗
Segment 3 · characters 531–765 · Original segment ↗
Segment 4 · characters 767–897 · Original segment ↗
Segment 5 · characters 899–960 · Original segment ↗
Quoted context
These are the quoted author’s words.
awaiting semi analysis post about whether this arizona tsmc thing is legit or mid
bayeslord ↗