I still don't understand why it's being done in Arizona. Sure, it's still economically worthwhile for the state government - but then it would also be viable in a state with better water resources, which would have a lower cost of water infrastructure, and so would be able to offer better tax incentives. What am I missing?

The Phoenix area has a long history of having semiconductor fabs starting with Motorola and Intel decades ago, so there is an established supplier base and workforce.

Makes sense, though that wasn't well explained in the article.

He also cited the seismic stability.

Because electricity there is cheap. That is because of the Hoover dam and cheap solar. Fabs use a lot of electricity.

Asianometry is quickly becoming one of my favorite tech information sources. Rapid, in-depth information that's very clearly explained. Always happy to see it showing up here.

I've spent many hours watching his videos. I wonder what his background is beyond this description:

Welcome! My name is Jon and I run the Asianometry YouTube channel.

I started the YouTube channel a few years ago. I moved to Asia after spending a decade in San Francisco, studying business. While the most popular videos seem to be about semiconductors and high technology, I do make videos about economics and history.

To be honest, I don't want to sound bad, but he seems to state the obvious thing which is available in Wikipedia or first page of google.

So, I don't think he provide rapid, in-depth information... . However, I do think he provide good information for layman like me. I have watched lithography, opensource related videos and they were general knowledge.

I agree; information discovery can be difficult.

I was so chilled, when I researched about water consumption by data center. They show us how datacenters etc are carbon neutral, but hide the water consumption. Semiconductor just don't have water problem, but many other issues like chemicals, high energy usage etc.. So, I think government should regulate company and enforce durable electronics item which should last minimum 10 years.

Datacenter water use is almost entirely for cooling. The water is evaporated into the air, and falls again as rain, effectively allowing that water to be reused.

In a large river basin, there is a very high chance the water falls again as rain into the very same river it was extracted from. In that case, it's a closed loop system.

This is unlike other water uses, like washing wafers, where the water will end up in a sewer and eventually into the sea.

The problem is when water evaporates and come as a rain, it is distributed. And, we need large amount of fresh water in one area. Additionally, you are not considering the time closed loop system takes to replenish the water source.

I agree with your last point. However, I think they still can filter and make it harmless, but we know there is no incentive to do that and the cost is high.

I recall desert places, particularly LV, treating and reusing their sewer water quite efficiently. In such places, the real water loss is whatever is sprayed onto the ground.

Can there be such a thing as a national water grid ala the electric grid? I'm curious if this has been fleshed out in any developed sense.

The desert southwest is truly huge and gets plenty of sun. So "just adding water" makes a lot of things viable.

This is a great article but 0 citations! It's super annoying to have to fact check every detail

eh it's pretty light on details really. Not much to cite. I think this particular article's best for his overall audience. Not meant to be a standalone policy brief.