Live data from Hacker News

Intel breaks ground on $20B Arizona plants as U.S. chip factory race heats up

reuters.com

181–190 of 326 posts

Re: Intel breaks ground on $20B Arizona plants as U.S. chip factory race heats up

#181

Earlier quoted context omitted.

The irony is that with current incentives, it's a self-reinforcing cycle. Nuts have a huge return on investment, so nut farmers can afford to buy even very expensive water, so they plant more trees, etc etc.

> nut farmers can afford to buy even very expensive water (x) doubt If they actually can pay for "very expensive water" i.e. water at the rates that residential/commercial users pay, then fair play to them. But from what I heard, they are only profitable by paying near-zero (orders of magnitudes less than other users).

What's the rough residential rate in AZ? Here in Seattle (obviously a MUCH different situation wrt water availability) I pay $6.96 per CCF (748.052 US gal, 3400 L). Even if every tree used an entire CCF over a season, that'd still be trivial compared to the amount of nuts you'd get (I assume)

Re: Intel breaks ground on $20B Arizona plants as U.S. chip factory race heats up

#182

What kind of ramp-up time are we looking at here? This can't be a quick process to get chips out the door I imagine!

I've heard that the general industry estimates for something like this are 5 (up to 10) years. However, both Intel and TSMC's USA are saying 2024. Potentially related is that 2024 is a presidential election year.

It's like trees: a long-term investment, and you might as well start now.

Re: Intel breaks ground on $20B Arizona plants as U.S. chip factory race heats up

#183

Earlier quoted context omitted.

Oh good, cities can be more water-efficient than people commonly believe. Guess it’s solved! Pay no attention to the drying rivers, emptying reservoirs, and depleting aquifers. If cities are more efficient per capita than people commonly believe, we can probably just grow our cities arbitrarily and our cotton farms and golf courses even more so.

Industrial and urban use cases can be very water resource friendly. Sure, maybe we shouldn't grow almonds in the desert. But that's not what's being discussed here.

My point is that saying "Arizona is among the best at driving itself into water depletion" is not a satisfactory answer to "Arizona is driving itself into water depletion."

Re: Intel breaks ground on $20B Arizona plants as U.S. chip factory race heats up

#184
post #5

Doesn't chip making a lot of water, most of it in Arizona is fed by Lake Mead? The same reservoir that is at 30% capacity, with acute water shortages and with 4 years of increased precipitation necessary for an adequate refill. We are adding a chip making facility which needs a lot of water here? This sounds like a recipe for disaster whenever there is a period of water stress which appears to be occurring in that re…

And yet you have the east coast(north east in my example) which is extremely lush and has 0 water issues and most companies are building these large sites in western water poor areas, intel - arizona phab, tesla - nevada gigafactory, facebook, datacenter southern new mexico.

Re: Intel breaks ground on $20B Arizona plants as U.S. chip factory race heats up

#185
post #181

Earlier quoted context omitted.

> nut farmers can afford to buy even very expensive water (x) doubt If they actually can pay for "very expensive water" i.e. water at the rates that residential/commercial users pay, then fair play to them. But from what I heard, they are only profitable by paying near-zero (orders of magnitudes less than other users).

What's the rough residential rate in AZ? Here in Seattle (obviously a MUCH different situation wrt water availability) I pay $6.96 per CCF (748.052 US gal, 3400 L). Even if every tree used an entire CCF over a season, that'd still be trivial compared to the amount of nuts you'd get (I assume)

California farmers use 40 trillion litres of water per year. At your rate that would be roughly 80 billion dollars a year or so. California's entire agriculture sector is worth around 50 billion a year dollars. So just the cost of water would be way more than all their earnings. And then you have to pay for equipment, pay for workers etc.

Re: Intel breaks ground on $20B Arizona plants as U.S. chip factory race heats up

#186
post #52

Earlier quoted context omitted.

> Seems to more closely follow tax breaks and tax income It follows feasibility, AFAIK. All new manufacturing is in Southern US states because they're the ones that allow new plants to be built. The others tie projects down in long, expensive, non-successful environmental studies. At this point, it's easier to bring water to the desert (Arizona) than get approval for a manufacturing facility in a non-desert.

It's the lack of unions and cheap wages. Southern states are famously lax on workers rights and business regulations.

Arizona is not a Southern state, and while it is a right-to-work state, the minimum wage there is higher than in most of the country.

Re: Intel breaks ground on $20B Arizona plants as U.S. chip factory race heats up

#188
post #5

Doesn't chip making a lot of water, most of it in Arizona is fed by Lake Mead? The same reservoir that is at 30% capacity, with acute water shortages and with 4 years of increased precipitation necessary for an adequate refill. We are adding a chip making facility which needs a lot of water here? This sounds like a recipe for disaster whenever there is a period of water stress which appears to be occurring in that re…

> Doesn't chip making a lot of water Can anyone share an insight on why this is the case? (I'm assuming it is true because I've heard the claim quite often on HN.) It's not really obvious to me where water is consumed during the fabrication process. It doesn't end up in the chips, so where does it go? If it were just used for cooling, it would be returned somewhere.

As mentioned, they are used to clean / wash off chemicals from wafers. And not just any water, but Ultrapure Water [1], it is an actual terminology, not marketing speak.

[1] https://en.wikipedia.org/wiki/Ultrapure_water

Re: Intel breaks ground on $20B Arizona plants as U.S. chip factory race heats up

#189

Earlier quoted context omitted.

So one business / structure will use 1/200th of all water? That does sound like a lot.

Not exactly. Fabs use a lot of water to run, but it's mostly self-contained with very little loss. I can't remember the exact number, but the fabs Intel already have in Phoenix only lose around 4-5% of the water. The rest is reclaimed and recirculated. I believe the two breaking ground today will be even more efficient, possibly even net producers once rain water and other sources are taken into account. It's been a…

net producers of water? check this please

Re: Intel breaks ground on $20B Arizona plants as U.S. chip factory race heats up

#190
post #144

Is there any evidence that Intel has fixed its process problems? Otherwise aren’t they building a $20B liability?

I think they're doing ok. I don't follow too closely, but 11th gen intel mobile is behind AMD. The good new for intel is they're not so far behind this generation that no one will buy their chips. Intel have new management and some optimism and by the time the plant is finished in 2-3 years they'll have their ducks in a row, but who knows what the market will be like then. In a bad scenario they could be a end up a b…

So many of the AMD laptops I see are geared towards gaming. Makes sense in that they're powerful, but I wish there were more options for general purpose lightweight laptops.
Post reply on HN