- cross-posted to:
- nova_scotia_general@lemmy.ca
- cross-posted to:
- nova_scotia_general@lemmy.ca
No portable gas turbines, no evaporative cooling. Grid connected with planned wind, closed loop cooling. Wow. So it IS possible to build these things without them being horror shows for the people around them. Assuming residential buildings are far enough for the noise to not be an issue.
Not saying it’s necessarily a good idea to build it, but if someone decided to spend the money anyway - it may not be a horror show.
EverWind said it will use a closed-loop water-glycol system. The company’s preliminary estimate for water usage is an average of 100,000 litres per day for the first phase and 500,000 litres of water per day including both phases.
A “closed-loop” that uses 100,000 litres per day doesn’t sound that closed.
As for the wind power, Everwind doesn’t appear to have even started construction on any of its planned wind farms and it’s likely the data centre will be completed well before they get any wind power online.
And this is all in service to an American company, completely throwing away the “sovereignty” justification for local data centres.
There’s nothing sovereign abt this for sure.
Yeah, even with that water usage it’s tiny compared to evaporative. Even 500K per day is on the very low end of industrial plants water usage, from what I gather.
E: Also I don’t know if this 500K liters disappeara like it does with evaporation or whether it’s returned for treatment and reuse. I’d assume if it isn’t horribly poisoned with something like PFAS, it would be returned to the lake via the treatment plant.
Too bad its dedicated to an American company, therefore subject to American intelligence rules, and Anthropic is also a company with deeply worrying financial projections. A main perpetrator of hype and bubble. Too bad we can’t support the development of Canadian or at least open-source models.
Be sure to get paid up front.
Fuck this crooked asshole and his liar crook business partners at Anthropic.
Unfortunate


