3 ms·

Somebody at Intel needs to play some serious hardball with market competitors and an internally ossified product development process. INTC has a deep patent po

Somebody at Intel needs to play some serious hardball with market competitors and an internally ossified product development process. INTC has a deep patent portfolio, incredible talent, fab capability and market access. Even moreso, it has capital market access. INTC could raise billions and billions for any endeavor in a short period of time. There are any number of scenarios and playbooks certainly gamed out by investment bankers over the years that Intel could pursue. What is lacking is a deep competitive drive driven by existential paranoia.

chx 6y ago
Billions and billions is nowhere near enough. https://www.eteknix.com/tsmc-building-fab-18-5nm-production/ https://www.eteknix.com/tsmc-building-fab-18-5nm-production/ > The entire cost of Fab 18 will cost TSMC $17.08 billion. Mind you, that's the 5nm production facility which is now complete and churning pretty good CPUs for Apple. They are also building a 3nm fab for 19.6 billion USD. Source https://www.taiwannews.com.tw/en/news/3805032 https://www.taiwannews.com.tw/en/news/3805032 To leapfrog TSMC, we are quite probably looking at something like 40-50 billion USD in R&D and fab costs and that's a gambit because they are now behind. It's one hell of a gambit for a 200B market cap company.
- AnimalMuppet 6y ago
  But is it more of a gamble than not doing it?
  - chx 6y ago
    I always had trouble finding wafer starts per months for Intel and TSMC both, if anyone have such data please share -- nonetheless, it seems like the demand for TSMC far outstrips supply and so there is plenty of room for Intel yet. In the worst case... there's a lot of money to be made as #2 chipmaker even if it's less than they are making as #1 chipmaker?
    - avs733 6y ago
      good luck, there is a reason you can't find it. I am not sure I have seen evidence for what you say. That is a complicated assertion because it speaks to wafer's started, wafer's finished, yield, die size, and technology among other things. Honestly, I suspect that Intel's struggles (as per typical) with new technologies are around ossification of technical paths, struggles with yield, and copy-stupid. That leaves them struggling to produce the quantity that customers need, and has opened avenues for customers to find other routes that provide them with more control. They are also seeing the benefits of custom silicon and the possibilities of full stack optimization, as opposed to optimization of just one or two parts which largely drove older approaches to software/hardware/battery efficiency. My other comment is that SPENDING on R&D is not always the best metric...
- mahkeiro 6y ago
  17 billion $ is a joke the Intel share buyback program is 20 billion $.... instead of buying their own share they can do that without raising or borrowing a single cent.
  - klelatti 6y ago
    But ... this article is a sign that investors are losing patience and so they might not have a free hand. It might seem straightforward to cancel buybacks but if you lose your investors then you might find its someone else implementing a new strategy.

chx 6y ago

Billions and billions is nowhere near enough. https://www.eteknix.com/tsmc-building-fab-18-5nm-production/ https://www.eteknix.com/tsmc-building-fab-18-5nm-production/ > The entire cost of Fab 18 will cost TSMC $17.08 billion. Mind you, that's the 5nm production facility which is now complete and churning pretty good CPUs for Apple. They are also building a 3nm fab for 19.6 billion USD. Source https://www.taiwannews.com.tw/en/news/3805032 https://www.taiwannews.com.tw/en/news/3805032 To leapfrog TSMC, we are quite probably looking at something like 40-50 billion USD in R&D and fab costs and that's a gambit because they are now behind. It's one hell of a gambit for a 200B market cap company.

AnimalMuppet 6y ago
But is it more of a gamble than not doing it?
- chx 6y ago
  I always had trouble finding wafer starts per months for Intel and TSMC both, if anyone have such data please share -- nonetheless, it seems like the demand for TSMC far outstrips supply and so there is plenty of room for Intel yet. In the worst case... there's a lot of money to be made as #2 chipmaker even if it's less than they are making as #1 chipmaker?
  - avs733 6y ago
    good luck, there is a reason you can't find it. I am not sure I have seen evidence for what you say. That is a complicated assertion because it speaks to wafer's started, wafer's finished, yield, die size, and technology among other things. Honestly, I suspect that Intel's struggles (as per typical) with new technologies are around ossification of technical paths, struggles with yield, and copy-stupid. That leaves them struggling to produce the quantity that customers need, and has opened avenues for customers to find other routes that provide them with more control. They are also seeing the benefits of custom silicon and the possibilities of full stack optimization, as opposed to optimization of just one or two parts which largely drove older approaches to software/hardware/battery efficiency. My other comment is that SPENDING on R&D is not always the best metric...
mahkeiro 6y ago
17 billion $ is a joke the Intel share buyback program is 20 billion $.... instead of buying their own share they can do that without raising or borrowing a single cent.
- klelatti 6y ago
  But ... this article is a sign that investors are losing patience and so they might not have a free hand. It might seem straightforward to cancel buybacks but if you lose your investors then you might find its someone else implementing a new strategy.