Smart to buy a preexisting fab. From what I understand via from various blogs and YouTube research, building a Feb even if your not doing cutting edge tech like tsmc, say going with an openpdk, still requires that special infrastructure. You need seismic dampening for the fab and to be located in a low activity region preferably, you need cheap water that can be refines, you need affordable electricity, and then supporting infrastructure to get the chemicals, the water, the machines delivered. Doing all of this isn't cheap and I'd bet is a lot of paperwork.

Imagine going to some rural area and trying to build a fab, chances are the town has no clue what your impact or needs are and you would be spending lots of money to basically speed up development of the area.

Side not to Semiconductors fab, where do you even buy one. Sure you can buy talent or machinery and then hire engineers to help get everything working, but if you wanted to for some reason buy a fan that already exists, say just the fab location and the equipment, how do you know what company to approach that might even consider selling. Who can even afford these purchases except massive fortune 500 company's breaking a piggy bank, or some massive credit institution, which I doubt would even do this because it would probably be a massive loan to any buyer. Seems like you need to have the money to build part of a fab if you want to buy one, idk who would even consider loaning that amount of money to a third party.

whats the process node it can do. im suprised a fab built in 1991 is stil valuable. the us is trying to build a 5nm fab in arizona and by the time its done well be at 1 or less. granted still probably worth it since it would only be about 7 years old process but 1991. thats gotta be old unless they kept upgrading it

Seems like it was previously owned by Coherent, like some kind of III-V (specifically GaAs mentioned) photonics processes there in the past. This kind of technology is typically quite useful for lasers, LEDs, or potentially image sensors as well. Many LIDAR sensors and even light sources can notably depend on III-V semiconductor sensors. Also widely used by the telecom industry.

Outside photonics definitely useful for high speed electronics, but that would probably take more process development to get going.

Why gallium arsenide? It's quite fussy.

"This facility is the only secure site in the UK capable of manufacturing gallium arsenide semiconductors, a vital component in military platforms such as fighter jets."

It’s hardly gonna be mass production with 100 staff, hardly worth mentioning. I think Motorola had a fab in the South West once upon a time. And Inmos in Newport in the South East.

Wasn't the last British owned semiconductor factory in the UK sold to Chinese investors within the last year it so.

The last steel furnace closing too?

This sure feels like yet another sign that major global powers are all gearing up for war.

This could be as benign as a government ensuring that the 100 jobs aren't lost, but given everything going in in both Europe and the Middle East it sure seems like more than saving such a comparatively small number of jobs. They could have just signed large(r) contracts with the company to financially secure the company, acquisition is a stronger play when the government needs more direct (and more private/secure) control.

Isn't it not how market works, by paying good price and allowing fabs to compete for it?

> The acquisition is expected to secure up to 100 skilled jobs in the North East and safeguard a critical part of the UK’s defence infrastructure.

The “North East” in the context of the UK and “the North East of England” are not interchangeable.

microchips aren't going to do much in a fire fight, they should probably re-industrialize

What lith node is that?

what a load of crap

'With global semiconductor demand rising, this move positions the UK to meet future technological needs, including advancements in artificial intelligence, quantum technologies, and 6G'?

advancements in artificial intelligence depend on mass production of 4nm silicon cmos, not 100 people doing gallium arsenide for high-speed analog. 'quantum technologies' is vague enough to not be literally a lie (transistors depend on quantum physics to work, as do wires) but in this context it's clearly designed to trick people into thinking 'quantum computing' which is also unrelated to what these guys are doing

so it wasnt to save them from insolvency?

Everyone who worked for Inmos rolls their eyes..

Here come the warm jets