View Single Post
      04-18-2024, 10:22 PM   #7903
x622
Forced Induction Connoisseur
x622's Avatar
980
Rep
758
Posts

 
Drives: 23 X5MC / 23 720s
Join Date: Aug 2020
Location: AZ

iTrader: (0)

Garage List
2024 G87 M2  [0.00]
2023 Mclaren 720s  [0.00]
2005 Honda S2000  [0.00]
2008 BMW M3  [0.00]
2023 BMW X5MC  [0.00]
1964 Ford Mustang  [0.00]
1968 Pontiac GTO  [0.00]
Quote:
Originally Posted by Dan B View Post
I apologize, I said dies. When you look at the failure rate across the same die it has gone down over time. And advances in manufatruring have even improved some numbers on smaller dies.

The point is, as manufacturing improves, costs typically go down.
You picked a pretty poor example;



The only thing that's really improved is old node yields, but that's already a sunk cost. Each node continues to get more and more expensive which is why fewer and fewer companies can afford get chips made for them in that process. Leading edge stuff will almost always go to the most cutting edge node, which typically has lower yields.

If you're trying to claim that reticle sizing has anything to do with it, that's not the case. Reticle sizing has stayed pretty much the same, wafer sizing is still basically the same.

Semiwiki and semianalysis are great resources if you'd like to learn more
__________________
Oy vey, look at all these shekels