GlobalFoundries Halts 7nm Development

[Trixanity]

Just saw that AMD did have a chip planned for Glofo but it wasn't supposed to tape-out till Q4 so that probably saved them a lot.

[Arika]

cool so now TSMC has a monopoly on 7nm

[Drak3]

18 minutes ago, Princess Cadence said:

Now where are all those Intel haters making jokes about 10nm at? two sides of the coin eh?

Get rekt AMD, the partner you don't use that much isn't following through with their 7nm that you planned on barely using at all.

[cj09beira]

19 minutes ago, Princess Cadence said:

Now where are all those Intel haters making jokes about 10nm at? two sides of the coin eh?

a filthy rich company not being able to deliver a 3 year old promise is very different from a "small" poor company running out of R&D money

[Dylanc1500]

6 minutes ago, Arika S said:

cool so now TSMC has a monopoly on 7nm

Not really, as "7nm" is more of marketing term than anything.

[Arika]

2 minutes ago, Dylanc1500 said:

Not really, as "7nm" is more of marketing term than anything.

they have a monopoly on manufacturing it since no one else is able to get down to it.

[Vode]

8 minutes ago, cj09beira said:

a filthy rich company not being able to deliver a 3 year old promise is very different from a "small" poor company running out of R&D money

Jup. Shouldn‘t have wasted billions on useless aquisitions like MacAfee.

[cj09beira]

2 minutes ago, Dylanc1500 said:

Not really, as "7nm" is more of marketing term than anything.

the only other fab that can make a chip with similar specs is samsung and they are late and almost don't do work for other companies

[cj09beira]

1 minute ago, Vode said:

Jup. Shouldn‘t have wasted billions on useless aquisitions like MacAfee.

thats true  

[S w a t s o n]

38 minutes ago, Princess Cadence said:

Now where are all those Intel haters making jokes about 10nm at? two sides of the coin eh?

10nm is an engineering failure from the so called "leading fab" in the world funded by the money printing machine named Intel. 7nm from GloFo was on track but their parent company doesn't want to bankroll more R&D because they aren't profitable. Not really the same problem

[Taf the Ghost]

58 minutes ago, Arika S said:

cool so now TSMC has a monopoly on 7nm

Samsung still exists. Nvidia actually makes the 1050 GPUs over there. We can generally expect AMD to look at some products on Samsung nodes over the next few years, I imagine.

 

In other news, where is IBM going to be moving production?

[Taf the Ghost]

Also, and this is kind of important, we might see AMD's entire 7nm portfolio a lot sooner than expected. GloFo was going to be Summer 2019, so Ryzen 3rd gen in April, maybe?

[yian88]

best news all year, i was afraid ryzen 3000 will be made with cheaper GF 7nm versus TSMC 7nm for EPYC

i only hope TSMC can produce high volume at 7nm for 2019 and beyond, otherwise we will see very expensive ryzen on zen 2 arch

 

i hope with all the people leaving RTG that they can deliver a decent navi GPU + gddr6, no morre HBM delay fails

[Taf the Ghost]

22 minutes ago, yian88 said:

best news all year, i was afraid ryzen 3000 will be made with cheaper GF 7nm versus TSMC 7nm for EPYC

i only hope TSMC can produce high volume at 7nm for 2019 and beyond, otherwise we will see very expensive ryzen on zen 2 arch

 

i hope with all the people leaving RTG that they can deliver a decent navi GPU + gddr6, no morre HBM delay fails

HBM3 at some point on 7nm, but that is for probably late 2019 with the Memory ramp. (DDR5 at the same time.)

[S w a t s o n]

46 minutes ago, yian88 said:

best news all year, i was afraid ryzen 3000 will be made with cheaper GF 7nm versus TSMC 7nm for EPYC

i only hope TSMC can produce high volume at 7nm for 2019 and beyond, otherwise we will see very expensive ryzen on zen 2 arch

 

i hope with all the people leaving RTG that they can deliver a decent navi GPU + gddr6, no morre HBM delay fails

GF7nm at least for AMD was basically TSMC 7nm, so I'm not sure it would have been demonstrably worse unless they fucked something up. TSMC is probably gonna accelerate plans to increase fab capacity at this point imo

[pas008]

Maybe they are skipping 7nm and going straight up 5nm

 

 

This article is over year old

 

https://arstechnica.com/gadgets/2017/06/ibm-5nm-chip/

 

Long shot speculation

[Space Reptile]

5 hours ago, TVwazhere said:

I just heard the news from my local news station. Hopped online and found this from the times union

 

https://www.timesunion.com/business/article/Layoffs-underway-at-GlobalFoundries-in-Malta-13011265.php

 

Looks like for all their efforts, TSMC is still kicking their asses and they're shifting to grab more market share. The article here goes on to say 7nm processes wont be affected, but everywhere else including my news station said the opposite. Mind you, My news station telling the public "GF will be halting 7nm chips and focusing on 12/14nm chips" to me was funny because I'm sure 99.5% of the people watching didnt know what that meant and dont care unless their jobs are on the line.

stopping 7nm and loosing AMD, one who would use that node in massive capacities, is not helping GoFlo to become more profitable and competitive 

[Space Reptile]

4 hours ago, Princess Cadence said:

Now where are all those Intel haters making jokes about 10nm at? two sides of the coin eh?

well TSMC / AMD still have working 7nm.... so the joke still stands 

[Taf the Ghost]

11 minutes ago, Space Reptile said:

stopping 7nm and loosing AMD, one who would use that node in massive capacities, is not helping GoFlo to become more profitable and competitive 

Depends how much it'll cost to take 7nm into High Volume production. If it's, say, 5 billion USD to retrofit Fab 8 in NY for 7nm, are they going to get the return out of that when their 14nm process is one of the best yielding in decades? They really needed a new Fab site, but there is also talk that yields with quad-patterning just aren't good. Without good yields, the huge volume products won't move to 7nm until it's far more cost effective.

 

New designs on 45nm, 28nm and, occasionally, 90nm pop up all the time. "Leading Edge" doesn't mean those are the only Fabs. Leading edge at high volume has dropped off a lot as it keeps taking more & more R&D costs. So companies license tech and spring up several years behind. This is the "China DRAM" market in a nutshell.

[leadeater]

10 minutes ago, Taf the Ghost said:

Depends how much it'll cost to take 7nm into High Volume production. If it's, say, 5 billion USD to retrofit Fab 8 in NY for 7nm, are they going to get the return out of that when their 14nm process is one of the best yielding in decades? They really needed a new Fab site, but there is also talk that yields with quad-patterning just aren't good. Without good yields, the huge volume products won't move to 7nm until it's far more cost effective.

 

New designs on 45nm, 28nm and, occasionally, 90nm pop up all the time. "Leading Edge" doesn't mean those are the only Fabs. Leading edge at high volume has dropped off a lot as it keeps taking more & more R&D costs. So companies license tech and spring up several years behind. This is the "China DRAM" market in a nutshell.

Not everything needs ultra density or ultra performance either, I can see why they would be keen on sticking with their high yield well refined 12nm and 14nm processes and just making customizations as required per customer. Easy money vs hard money, I'd take easy money myself.

[Taf the Ghost]

4 minutes ago, leadeater said:

Not everything needs ultra density or ultra performance either, I can see why they would be keen on sticking with their high yield well refined 12nm and 14nm processes and just making customizations as required per customer. Easy money vs hard money, I'd take easy money myself.

Considering they haven't been making a lot of money, the logic makes sense. AMD was quite well hedged for this.

 

The other little wrinkle is "did they need EUV first?". If they get their EUV machines in over the next few years, they might jump back into 5nm earlier than you'd expect.

[GoldenLag]

1 hour ago, VegetableStu said:

ZEN++

Nah, ZEN2-

[S w a t s o n]

1 hour ago, leadeater said:

Not everything needs ultra density or ultra performance either, I can see why they would be keen on sticking with their high yield well refined 12nm and 14nm processes and just making customizations as required per customer. Easy money vs hard money, I'd take easy money myself.

 

1 hour ago, Taf the Ghost said:

Considering they haven't been making a lot of money, the logic makes sense. AMD was quite well hedged for this.

 

The other little wrinkle is "did they need EUV first?". If they get their EUV machines in over the next few years, they might jump back into 5nm earlier than you'd expect.

Honestly, I feel like they cut their losses too early. 1st gen 7nm was so close to being ready and they basically just burned all that money spent on development. They could have at least revised the roadmap to be much slower or something, the volume AMD/IBM would have provided probably could have made it viable to run with 1st gen for awhile

[Taf the Ghost]

20 minutes ago, S w a t s o n said:

 

Honestly, I feel like they cut their losses too early. 1st gen 7nm was so close to being ready and they basically just burned all that money spent on development. They could have at least revised the roadmap to be much slower or something, the volume AMD/IBM would have provided probably could have made it viable to run with 1st gen for awhile

R&D covers a lot. You can do all of the design work for a multi-billion dollar project, but the big cost is when you go to start physically building it. The big bill comes when you have to build the fab to move to high volume. Some of the scuttlebutt is that they'd have needed to do it at Fab 8 in NY, which probably meant no more 14nm customers. Though it really just leaves us with a lot more questions than answers.

 

Is IBM moving somewhere else?

Are they still taking delivery on EUV machines?

Are they going to show back up with 5nm in 2022?

Will they continue the research Fab until they can license the technology to trailing Fabs?

Can their research Fab handle IBM's volume?

Are they just going to license future nodes anyway?

 

HVM & research was canceled, but 7nm at GloFo is probably not dead. So, we'll just have to wait and see. These things have weird ways of playing out in tech.

[Stefan Payne]

8 hours ago, Princess Cadence said:

Now where are all those Intel haters making jokes about 10nm at? two sides of the coin eh?

Ever heard about TSMC??

You know that Samsung also does Chip manufacturing?
 

AMD doesn't need GF, never did.

GF always needed AMD.

 

See the Problem?

 

cool so now TSMC has a monopoly on 7nm

No, there is still SAMSUNG.