https://www.anandtech.com/show/21266/globalfoundries-clients-are-migrating-to-sub10nm-faster-than-expected
13 replies (most recent on top)
AMD was wrong to build a fab in NY but right to cut bait and get away from the foundry business. Production foundry's need to be in Asia and smaller development in USA.
Apple figured this out long ago, designed in Cupertino, built in China.
Some very to the point comments here. I'd like to add that the entire idea of building a fab in a high cost area to compete with lower cost Asian suppliers was doomed. I don't know what the investors were thinking. State and federal taxpayer money can only bridge the gap for so long. Our so called leadership is not very bright but even they could see that, they were just hoping to cash out before everything comes crashing down.
They were always lying to us. The entire business model is not viable.
Personally, I'm really pi---d off about how much blood and sweat I've put into GF and how TC and his minions have run this company into the ground. TC needs to go and his SLT minions need to follow him out the door. Let's re-shuffle the deck with new SLT leaders (actually let's get a whole new deck of cards) who actually have a brain and can come up with a winning strategy which leads to differentiated technologies and industry leadership. GF1.0 under CEO SJ was a creative environment to work in but a financial mess. GF2.0 under CEO TC started out being financially responsible but the SLT su-ked so many millions out of the company and lost the majority of the talent due to cost cutting which led to having no differentiated technologies. Now the question is what will GF3.0 look like, and will they clean house and start over with a new CEO and completely upgraded SLT?
The talent has left the building. Only ones left are waiting for the package and July exit.
Couldn’t be, ol’ Tommy just told us how our customers are begging him to save them from single digit nanometer…
Already rumors of more resource actions at fab9. Even after the annual December layoffs.
I'm just waiting for a package, I now work accordingly to how I am treated.
It will take tools and talent ! GF refuses to invest in either.
This was always Globalfoundries fate. The trajectory they put us on during the pivot it was inevitably this. And all those who had hope or were tricked into thinking this was a good business decision are truly naive mo--ns. The pivot was a shortsighted interim band-aide fix for our financial situation at that point in time. The problem is our leadership didn’t work and plan for the future years out. It should have been everyone’s question at the time; what do we do when the bulk of customers migrate to more advanced nodes? Everyone should have pressed this point. The only response we really got was there will always be customers that use these older mature nodes. But reality is those customers also migrate to what has become the ‘new’ old mature node. Leaving us standing there with our di-ks in our hands.
The pivot could have worked if a strategy and a proper roadmap plan had been laid out for the future of our business objectives. Some did asked for it, but it was always dismissed or told it will be shown later. Well a true plan never materialized just vague bullsh-t. So we became reactive to business situations and environments. We have been on our heels ever since and it’s finally catching up to us. We became a company who improvises on the spot and more often than not incorrectly. We should have had better planning and working towards some future position.
Our leaders have fumbled and tripped over their own feet on nearly every major business decision that has been presented to them. And the fruits of those labors are coming to bare.
They had the EUV tools installed, and sold them at a garage sale.
Soon we will be pivoting to 200mm and 92nm tech. It's the future™
The end of Fab 9 anyway.
Pivoting to 7nm and below would require EUV tools, baseline tool around $250M and the newest EUV tool around $450M. To justify that investment there would need to be enough demand in those nodes. Then even if GF buys EUV tools they would have to ramp up the whole ecosystem around them as well which could take many years. If GF wanted to go 7nm and below it would take 3+ years or probably more to even get EUV up and running and then buy a tech and transfer another couple years. Net, 5+ years for GF to get a 7nm or below tech qualified. I don't believe the business case is there. Now the big problem is TC sold everyone that 70% of the market would be in 12nm and older through at least 2030 but it seems that is changing so GF's TAM and SAM numbers will need to be modified to expect a decline in overall revenue moving into the future unless they take more market share which won't happen given their non-feature rich technologies. I don't see GF going bankrupt or losing money, but it will be a bottom feeder foundry falling further behind UMC and potentially only competing with Tower-Jazz for the scraps left over which TSMC, Samsung, Intel and UMC don't take. When one is a bottom feeder, it will be all about costs, so prepare for more low-cost geo outsourcing and lagging industry salaries. Some might not like this message/post but it is what will happen moving into the future so be prepared.
The end? Nah, no way. GF can just... ahhh...
PIVOT! Yeah, that's it. They can pivot and go back to 7nm and beyond. It's not too late, right?