News GlobalFoundries and IBM have buried the hatchet, paving the way for future collaboration

It's hard to understand why Global Foundries gave up on 7nm in light of this. AMD had suddenly turned around and its sales were growing rapidly, and not delivering 7nm meant that Global Foundries missed out on a lot of that success. Even if it had been the last new node, GloFo 7nm would probably have been the I/O die for Ryzen 9000 desktop chips. Global Foundaries missed out on that and exposed itself to liability from IBM. I can't imagine why unless AMD was already interested in leaving for TSMC.
 
  • Like
Reactions: gg83
It's hard to understand why Global Foundries gave up on 7nm in light of this. AMD had suddenly turned around and its sales were growing rapidly, and not delivering 7nm meant that Global Foundries missed out on a lot of that success. Even if it had been the last new node, GloFo 7nm would probably have been the I/O die for Ryzen 9000 desktop chips. Global Foundaries missed out on that and exposed itself to liability from IBM. I can't imagine why unless AMD was already interested in leaving for TSMC.
AMD sold their fabs because they couldn't generate enough revenue to operate leading edge with their volume. Apple was firmly entrenched with TSMC, and there were no other big volume leading edge players. GloFo canceled development of their 7nm process the year TSMC was shipping products on it which should also indicate how far behind they were with development. While I agree the company's long term health would have been better it was hemorrhaging money still and not investing in EUV machines was the way they chose to stem the bleeding.
 
  • Like
Reactions: gg83
It's hard to understand why Global Foundries gave up on 7nm in light of this. AMD had suddenly turned around and its sales were growing rapidly, and not delivering 7nm meant that Global Foundries missed out on a lot of that success. Even if it had been the last new node, GloFo 7nm would probably have been the I/O die for Ryzen 9000 desktop chips. Global Foundaries missed out on that and exposed itself to liability from IBM. I can't imagine why unless AMD was already interested in leaving for TSMC.
You have to have customers to justify the investment into leading edge. EUVs, fabs, equipment are very costly, you need revenue to offset and recoup the huge investment. Intel is experiencing this dilemma as we speak, almost no major secured customers, definitely not enough to justify the costly investments into fab expansion and 18A.
 
Last edited:
  • Like
Reactions: gg83
I hope GF can work with IBM to produce future tech here in upstate New York. The timing really sucked for the 7nm chips. AMD probably couldn't predict how many chips they would need. If AMD and IBM were the only ones at the time interested in GF 7nm chips, that wouldn't be enough. They were having terrible yields going smaller than 12nm so they had to stop. But then budget Asian fabs are making legacy nodes, so what does GF do now? American military contracts i guess?