IBM suing GlobalFoundries over contract dispute

Lakados

[H]F Junkie
Joined
Feb 3, 2014
Messages
10,379
https://www.reuters.com/technology/globalfoundries-seeks-ruling-ibm-contract-dispute-2021-06-08/
https://www.extremetech.com/computi...or-2-5b-over-wrecked-process-node-transitions

TLDR;
IBM partnered with GF for chips to be built on 10 and 7 nm
IBM gave GF access to patents and money to upgrade the processes to make the chips
GF didn't deliver, and instead canceled their 10 and 7nm plans
IBM forced to partner with Samsung to have chips made instead
IBM wants their money and patents back

Edit:
GF's defense is apparently "Samsung did it cheaper and faster than we could have anyways, so really we did IBM a favor."
 
Last edited:
IBM is probably one of the last companies i'd want suing me, good luck GF you are screwed!
 
New Headline, "IBM acquires Global Foundries", lol
That would be the best outcome. IBM should never have gotten out of fabrication. Its crazy they are still doing fabrication R&D... with no production capacity of their own... who ever the exec was that thought that was a good idea I hope got turfed. IBM still has a industry competitive arch and chip designs. ARM is going to stroll in completely replace power in the server markets. Selling their fab was a big mistake for IBMs hardware division.
 
That would be the best outcome. IBM should never have gotten out of fabrication. Its crazy they are still doing fabrication R&D... with no production capacity of their own... who ever the exec was that thought that was a good idea I hope got turfed. IBM still has a industry competitive arch and chip designs. ARM is going to stroll in completely replace power in the server markets. Selling their fab was a big mistake for IBMs hardware division.
Yes and no they sold it to GF with a 10-year agreement where they were going to be the sole provider of chips to IBM, IBM was doing the R&D, GF was going to handle the manufacturing. IBM doesn't need enough first-party chips to warrant their own fabs anymore, so the original plan was for IBM to get GF the tech they need to make all their chips, and then they could sell off any leftover fab time. But a few years in it seems like GF decided that IBM wasn't ordering enough chips and just let the whole agreement slide, then cancelled any further upgrades.

This is all coming out now because it looks like GF is about to offer up an IPO and try to go public.
 
Yes and no they sold it to GF with a 10-year agreement where they were going to be the sole provider of chips to IBM, IBM was doing the R&D, GF was going to handle the manufacturing. IBM doesn't need enough first-party chips to warrant their own fabs anymore, so the original plan was for IBM to get GF the tech they need to make all their chips, and then they could sell off any leftover fab time. But a few years in it seems like GF decided that IBM wasn't ordering enough chips and just let the whole agreement slide, then cancelled any further upgrades.

This is all coming out now because it looks like GF is about to offer up an IPO and try to go public.
Good luck to them. I never hear any good news about them lol
 
That would be the best outcome. IBM should never have gotten out of fabrication. Its crazy they are still doing fabrication R&D... with no production capacity of their own... who ever the exec was that thought that was a good idea I hope got turfed. IBM still has a industry competitive arch and chip designs. ARM is going to stroll in completely replace power in the server markets. Selling their fab was a big mistake for IBMs hardware division.

you really dont know anything about the ceo's at ibm.
ketchup boy most certainly did NOT get "turfed" and ms mistake not only did not get "turfed" but split the company in 2 so she could be on both boards and get 2 golden parachutes. p boy didnt do the company any favors either. green acers was the 1st ceo to begin layoffs.
so in all told. the last 4 ceo's - going back to '85 - havent helped the company much (mainfraime not withstanding and then its demise) except to raise the share price to make their solid gold parachute so very nice
 
  • Like
Reactions: ChadD
like this
IBM gave GF no where near enough to transition to 7nm. Also the time tables are absurd. I think the deal started in 2015. So in 5 years they where going to go to 7nm? GF? Yeah right. For 200 million? Lol
 
IBM gave GF no where near enough to transition to 7nm. Also the time tables are absurd. I think the deal started in 2015. So in 5 years they where going to go to 7nm? GF? Yeah right. For 200 million? Lol
Well GF agreed and thought it was a reasonable request at the time. If they had continued to at least work on advancing their nodes at all then things probably would have worked out, but throwing their hands up at 14nm saying "we're good now boys" and looking to go public with an IPO offering was probably the last straw for IBM.
 
What parents does IBM have that could be useful? They aren't exactly chip builders are they?
 
What parents does IBM have that could be useful? They aren't exactly chip builders are they?
They made the Cell archetecture which was used in the PS3 and is used in some supercomputers, as well as many other computing technologies.
 
Well GF agreed and thought it was a reasonable request at the time. If they had continued to at least work on advancing their nodes at all then things probably would have worked out, but throwing their hands up at 14nm saying "we're good now boys" and looking to go public with an IPO offering was probably the last straw for IBM.
But everyone knew they weren't going to 7nm even IBM. That 200 million had to be for something else because going to 7nm was expensive. GF was not going to get there before Intel for 200 million. Lol ...I don't think so. That amount of money sounds more like the cost of an order.
 
But everyone knew they weren't going to 7nm even IBM. That 200 million had to be for something else because going to 7nm was expensive. GF was not going to get there before Intel for 200 million. Lol ...I don't think so. That amount of money sounds more like the cost of an order.
Well, yeah. IBM isn't going to pay for a node transition unless they're guaranteed a metric ton of wafers, and it would surely be more than $1B
 
Back
Top