When it comes to chip foundries, most the value in IP is from working and commercially-tested processes, not things dreamt up and never made reality - any physics grad can do that - because you don't know how much effort it'll take to make it work. This is particularly poignant for GloFo, who abandoned leading-edge work and forced IBM to seek alternative partners.
The problem for IBM is that things dreamt up decades ago are now being made reality, and those dreams are the property of GloFo. IBM is totally fucked here if a court and counsel savvy enough to explain the correlations is assigned.
As far as IBM asking for 2.5B in damages from GloFo over shifting roadmaps...that's even more of a stretch and probably spawned this lawsuit so there can be a settlement.
"newly formed leading-edge logic foundry backed by leading Japanese companies"
Seriously? Way to bury the lede!
The very existence of Rapidus seems rather more important than this minor legal squabble. And the fact that I have heard nothing about it before (as someone who follows this space not obsessively, but more so than the average person) even in these times of obsessing over Intel's foundry and what might happen to TSMC suggests that the news media have seriously let us all down!
How leading edge? Wikipedia says they hope to have 2nm by 2027, so just two or three years behind TSMC! (Essentially 2nm is code for GAA transistors on high-NA EUV -- omit either of those and you are talking PR hype, not leading edge.)
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
10 Comments
Back to Article
shabby - Thursday, April 20, 2023 - link
Global foundries has leading edge chip ip? 😂LordConrad - Thursday, April 20, 2023 - link
Completely possible. Having IP and putting it to use are two very different things.Threska - Thursday, April 20, 2023 - link
AMD use to use them for the I/O chiplets.meacupla - Thursday, April 20, 2023 - link
Yeah, and the X570 chipset that gloflo made ran hot, and everyone was making fun of AMD for requiring a 40mm fan to cool it.The only way that's "cutting edge" would be if the fan chopped fingers off.
Threska - Thursday, April 20, 2023 - link
Fixed with the X570S.Wrs - Thursday, April 20, 2023 - link
When it comes to chip foundries, most the value in IP is from working and commercially-tested processes, not things dreamt up and never made reality - any physics grad can do that - because you don't know how much effort it'll take to make it work. This is particularly poignant for GloFo, who abandoned leading-edge work and forced IBM to seek alternative partners.Samus - Friday, April 21, 2023 - link
The problem for IBM is that things dreamt up decades ago are now being made reality, and those dreams are the property of GloFo. IBM is totally fucked here if a court and counsel savvy enough to explain the correlations is assigned.As far as IBM asking for 2.5B in damages from GloFo over shifting roadmaps...that's even more of a stretch and probably spawned this lawsuit so there can be a settlement.
erinadreno - Friday, April 21, 2023 - link
Your car's MCU probably runs on GF's IP. And so does your AC, monitor, and your phone chargername99 - Thursday, April 20, 2023 - link
"newly formed leading-edge logic foundry backed by leading Japanese companies"Seriously? Way to bury the lede!
The very existence of Rapidus seems rather more important than this minor legal squabble.
And the fact that I have heard nothing about it before (as someone who follows this space not obsessively, but more so than the average person) even in these times of obsessing over Intel's foundry and what might happen to TSMC suggests that the news media have seriously let us all down!
How leading edge? Wikipedia says they hope to have 2nm by 2027, so just two or three years behind TSMC! (Essentially 2nm is code for GAA transistors on high-NA EUV -- omit either of those and you are talking PR hype, not leading edge.)
Threska - Thursday, April 20, 2023 - link
IBM use to be leading edge research. We'll see if they can translate that to something cutting-edge.