https://www.nextplatform.com/2021/06/10/why-ibm-is-suing-globalfoundries-over-chip-roadmap-failures/
6 replies (most recent on top)
1nm is announced .. not from IBM but from the world's largest chip manufacturer ...
https://hothardware.com/news/tsmc-and-mit-research-team-claim-1nm-chip-breakthrough
"Think about it. You walk into a video store, you see 8-Minute Abs sittin' there, there's 7-Minute Abs right beside it. Which one are you gonna pick, man?
I would go for the 7."
Because they can - and struggle to make money with current operations!
IBM's problem is (still) decades later summed up by this old quote:
IBM : You can buy better, but you can't pay more.
IBM had an approx 18-24 month lead in chip technology. The National labs wins proved that. With GF exiting 10 and then 7 NM technology, they squandered IBM’s lead as it took Samsung that long to come up to speed. Shame on systems for not anticipating, but as a previous poster said, with less than 1% of chips sets worldwide, power most likely had to single source just to get a decent price per chip set. Now that the Power chip lead has been squandered, IBM is most likely repositioning it’s go to market strategy. The low end without the chip lead is most likely doomed as Intel has filled that vacuum. The high end with its next generation memory technology, most likely will find a NICHE market and continue on. I believe the high end chips will also be used in system z and high end storage to give those products a competitive advantage so again a reason to pursue the high end. Power has an interesting revenue mix when it comes to power. Approx 20% comes from Storage and non-power server installs. The remaining revenue is equally split high end vs low end. The high end is represented by approx 5% of all boxes made, with the low end capturing 95% of the boxes made. So if IBM concentrates on the high end, they capture 50% of the revenue for approx 5% of the boxes made. YES the high end boxes are larger so some extra effort and components need to be sourced, but you can see the profit margin difference high end vs low end.
Assume this is fiction:
IBM R&D can write a ton of patents and get one chip with a heartbeat out of a few hundred wafer starts. Nothing is scalable and GF got screwed. GF backs out.
IBM has better lawyers and are making it look like GF is the bad guy.
Even the new 2 nm is not really 2 nm. IBM just called it that. Since when is a nanometer not a set measurement. Again , good lawyers.