• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

GlobalFoundries is putting its 7nm finFET program on hold

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

wingman99

Member
Joined
Dec 10, 2003
Looks like Intel is not the only company having production cost problems with a node shrink. AMD, a big customer for GlobalFoundries, is moving its 7nm production to TSMC. We will see how TSMC will do with 7nm.


GlobalFoundries is putting its 7nm finFET program on hold indefinitely and has dropped plans to pursue technology nodes beyond 7nm. https://semiengineering.com/gf-puts-7nm-on-hold/
 
I think this was discussed somewhere last week or so...or the day that came out?

AMD moved to TSMC a few months ago due to this (and other) issues, that I do know.



EDIT: Found it. See what vague thread titles (like the one below) will do people? Make things so they cannot be found. :)

https://www.overclockers.com/forums...t-impact-do-you-think-this-might-have-for-AMD

I may merge the two threads and correct the title of the orignal.
 
on topic, have they reached a point that making them any smaller is just going to cost to much with current tech?
 
They are all having a low yield problems with the new 10-7nm EUV technology. GlobalFoundries is sitting this out due to cost of 7nm production.
 
for gf its already hit that point. i dont think its a good thing that they just up and gave up the 7nm ghost. but tsmc is already testing their amd 7nm cpus with success according to the latest news from them. so maybe amd will continue to ride their newfound success right into 5nm. apparently tsmc have already designated 25 BILLION cough cough hack(its hard to even type that kind of number) into 5nm. from the little i have read about it tho it isnt going to benefit desktop parts yet. just search tsmc and 5nm. youll find a slew of articles about amds plans. this is one i read a while back but is still relevant.
https://www.anandtech.com/show/1272...-scaling-but-thin-power-and-performance-gains
from what ive gleaned from reading about tsmcs surge is that 3.5nm is about as far as this process can go before it gets to expensive to support.

for intel the difficulty they have had simply going to 10nm/10nm+(not sure which anymore) makes it seem like they are waaaay behind the curve. they will just keep re-releasing slightly faster(cough) optimized 4, 6 and 8 core procs with a bit better efficiency(5.5-5.7 single on water maybe?).
i honestly cant say what is up with intel and their 10nm anymore there have been so many release dates i have given up trying to keep them straight. maybe q1/2 19 iuno. as deep as intels pockets are who knows when they will hit the nm wall. from what weve seen its going to be 2025 before we know anything for sure haha.
 
is there really a good reason to go so small?
I can't see that a larger die would bother anybody, my threaddrippers are huge compared to todays cpus.
 
More transistors closer together=less voltage drop between gates, more efficient, less time to transit making it faster. At least that's what I think the logic is.
 
does the 7-14 nm refer to the size of the transistor or the wire that connects them?
 
7nm are referring to the transistor node. However, Intel 10nm and TSMC 7nm use different measurements for the components of the transistor.
 
Back