Samsung to Build $15,000,000,000 Semiconductor Plant in Texas

Joined
Jan 31, 2006
Messages
4,440
Location
Idaho

The facility expands the South Korean tech giant's presence in Texas, which is currently limited to a chip plant in Austin. Samsung's new fab will be dedicated to the manufacture of silicon to be applied in a variety of segments including mobile, 5G, high-performance computing and artificial intelligence.

Taylor reportedly offered tax incentives to woo Samsung, including property tax breaks that equate to about 92.5% for the first 10 years, the WSJ reported Monday.

Biden and Congress are looking to bolster U.S. chip manufacturing through funding and industry subsidies. In June, the Senate approved $52 billion in subsidies for new chipmaking facilities, though the legislation has yet to be approved by the House.

The company plans to break ground on the 5 million-square-meter site in the first half of 2022, with hopes to have the facility up and running in the second half of 2024.
 
My grandparents were from Taylor … went up there many times when I was young …
It used to seem remote & ”country” … guess many places did way back when 😷
 
I hope they watch them closely.They've been caught doing some shady things in their factories.
 
I hope they watch them closely.They've been caught doing some shady things in their factories.
As in what?

I'm actually more concerned about the stability of the electric grid. If the "cold spell" happens again that'll be easily 100M lost to the factory. I hope they build their own power plant with all the weather proofing they need to protect themselves.
 
Glad to see the change is attitude in high tech manufacturing. We ignorantly gave the business away years ago.
“Potato chips, computer chips, what’s the difference? A hundred dollars of one or a hundred dollars of the other is still a hundred dollars.”

TSMC is building a huge fab in Arizona and now Samsung in Texas. Great news. It seems like we are taking high tech manufacturing seriously. Finally...
Each fab will require $10B or more in chip fab tools; manufacture of these tools is still dominated by America.
Applied Materials, Lam Research and KLA-Tencor are the major players. All Silicon Valley!
 
Glad to see the change is attitude in high tech manufacturing. We ignorantly gave the business away years ago.
“Potato chips, computer chips, what’s the difference? A hundred dollars of one or a hundred dollars of the other is still a hundred dollars.”

TSMC is building a huge fab in Arizona and now Samsung in Texas. Great news. It seems like we are taking high tech manufacturing seriously. Finally...
Each fab will require $10B or more in chip fab tools; manufacture of these tools is still dominated by America.
Applied Materials, Lam Research and KLA-Tencor are the major players. All Silicon Valley!

I worked at a company that still had a small production fab in Silicon Valley. I didn't necessarily have a whole lot of reason to go to that part of the campus, but once I was helping to debug something and went into the e-beam lab that was attached to the fab. In grad school one of my instructors worked at HP's fab and also an instructor at Stanford. I had to turn in some work late and arranged to drop it off at his desk at Stanford. It was really weird because it was an office building with a small research fab inside the building. Just office buildings but the fab inside of glass. He also took us on a weekend tour of the HP fab in Palo Alto where he worked. It was really weird seeing the tubing there that had signs warning of arsenic gas.

Fabs have never really left the US, although many of the big American players have definitely left the market, like AMD, IBM, HP, Lucent, etc. Intel has always relied on its American fabs, but for years they never took outside customers. But certainly Samsung isn't under American control.

I'll just say I've heard of some really weird things too, including major semiconductor companies that go to fabs like TSMC to make digital ICs. They really can't make those in their own fabs because their processes were developed for memories. But the industry is filled with a lot of things that seem contradictory, including that ARM's biggest customer is Intel.

It's primarily TSMC and UMC that dominates the market because they're really good at what they do. It's not really all that different than Pabst Blue Ribbon not actually making any of the beer they sell. They've got the recipe and Miller makes it for them.
 
Once upon a time the risk of finding someone else to build your stuff and fail is higher, then it change and TSMC started taking risk when nobody else would, and it became clear that it makes more sense to ask foundry to build your stuff than risk going out of business building your own and fail (3dfx).

The only remaining survivors are those with big enough volumes (Intel for CPU, and everyone else doing DRAM and NAND, the commodity stuff).

IBM could have done ok entering the foundry business, but the legacy MBA decided it is more profitable to sell the fab than going into a foundry business, plus a bunch of other businesses they were good at but weren't making enough profit to keep on the book. Intel just went downhill when they decided to move to low cost of living area and sit on the monopoly, paying low salary to their people. Having TSMC enter Intel's home turf and poach their engineers away is good, we need competitors to make the Intel more competitive in the long run, and their people better paid.

So what is the conclusion? Finance guys aren't great tech CEOs. I have great expectation of the new Intel CEO.
 
  • Like
Reactions: GON
So what is the conclusion? Finance guys aren't great tech CEOs. I have great expectation of the new Intel CEO.
I agree, but would say it differently. The CEO reports to the BOD and shareholders. His/her job is to maximize the wealth of the shareholders.
2nd, different leadership is needed at different times. Start up, growth, rapid growth, downturn, diversification, consolidation and more.
I worked for "the best run $500M company in the world". But that leadership did not scale much past $1B.
As a shareholder, different CEOs (3 in total over 20+ years) I was rewarded very well. And I am looking forward to more.

Finally, SEMI is thought by many to be the most demanding, hardest, most cut-throat industry in the world. No thin skin in our indsutry. Very few winners and TONS of losers.
 
Once upon a time the risk of finding someone else to build your stuff and fail is higher, then it change and TSMC started taking risk when nobody else would, and it became clear that it makes more sense to ask foundry to build your stuff than risk going out of business building your own and fail (3dfx).

I entered the industry around the time when "fabless" was a huge catchphrase. But back then TSMC was an option but certainly not the 900 lb gorilla that it is today. The companies I worked for initially used spare capacity at companies that mostly produced their own products.

Some of the "fabless" companies decided that they would get into their own fabs because they didn't necessarily get priority. The one company I was thinking of was Cirrus Logic, which had a couple of rather disastrous joint ventures including Micrus with IBM and Cirent with Lucent.

Cirrus Logic Inc. has returned to its roots as a fabless semiconductor company.​
The Fremont, Calif., chip maker-one of the pioneers of the fabless semiconductor business model-last week ceded to Lucent Technologies Inc. its interest in Cirent Semiconductors, their joint-venture wafer fab in Orlando, Fla. The agreement is expected to close by the end of the month, pending regulatory review.​
That move came on the heels of a similar accord that Cirrus Logic reached with IBM Microelectronics. Earlier last month, IBM agreed to assume full control of the companies' MiCRUS joint venture in East Fishkill, N.Y.​
 
My grandparents were from Taylor … went up there many times when I was young …
It used to seem remote & ”country” … guess many places did way back when 😷

I have family sprinkled all over East Texas (Taylor, Kilgore, Longview, Canton, Athens, & New Boston), Nice to visit but wouldn't want to live there.
 
  • Like
Reactions: 4WD
Back
Top