Intel
Scenarios Update August 2024 Edition (POST EARNINGS)
When
Intel announced IDM2.0 and many fabs in 2021, We highlighted a number of
issues:
- Intel is not a cost effective manufacturer. They are not agile and the wafer costs are 2x competition (it’s really easy math now that Intel has reported it out). They are focusing on their worst area. This is why the decision was made before Pat to outsource Fab work.
- Intel cannot afford to spend the capital. Revenue is shrinking, market share is shrinking. You cannot build fabs in hope that people buy.
- Intel needs to focus on what they do well. Leading compute architecture standard and products. Hold on to >50% share in DC CPU and PC CPU. Intel is not a good graphics chip, mobile or memory company
Q2
Earning report out:
Intel
revenue did not crash. It was flat. Intel has been losing money for some time
so that wasn’t different (GAAP earnings). Intel sold a lot of CPUs. What
became clear at the earning call was
- Intel is not growing in one of the fastest growth periods in compute. Negative DCAI growth
- Intel has no meaningful foundry business today. $77m in external revenue including packaging. They probably are not a top 20 external wafer foundry today
- Intel loses money on old processes and now when it ramps new processes. Foundry group losses are growing very fast. This despite the risky depreciation timeline that helps short term, hurts long term.
- Intel ship released products and claimed to be on track for 5N4Y. Apparently great nodes . But it doesn’t matter if you don’t sell them cost effectively.
As a result, they will cut costs until they can show a financially stable plan. Where will they cut? Will it work? what about those scenarios for foundry we published Which one is coming true
See our website for more details and Intel wafer capacity through 2030.
We will be at FMS in Santa Clara Aug 5-8. Text or email to set up a meeting
Mark Webb
