Is the “used” water by air data centers evaporated or flushed back into river/ lake/resevoir? And is it dirty?
People who live within 1 mile of a golf course are x times more likely to have cancer. So even though a portion of the water used on golf courses is run off, it’s heavily contaminated with chemicals.
That data about local temp increases is astonishing. Are we sure the study’s authors are correct? And how does it compare to other industries? A 9 degree celsius increase is catastrophic
Exactly. Forget climate change modelling then and revert to mapping locations of data centres and tracking extreme weather patterns locally.. if only it was that easy
Using 2023 #s for data center water cooling is substantially out of date, right?
Also, just because golf courses do it 30x more, it’s ok for data centers to guzzle 17B gallons of water, which presumably will double to 35B+ in short order, many times in areas where water table is already at the tipping point, eg Google DC in Dalles, Oregon ? Not convinced by this line of reasoning. At all.
Yes, I appreciate that we already waste water elsewhere. But a) averages have little meaning here, and b) bad is still, uh, bad. ;) When it’s your small-town watershed and Google moves in to consume an extra 1 million gallons per month, quadrupling water use and depleting the water table 10x faster than otherwise (draining the last creek), talking about almonds doesn’t help solve the problem. Nor does using more of an overused resource make it less bad because someone else overused it first. But I’m all for taxing golf courses and AI inference to pay for watershed restoration ;)
That i think is the point. first of all water use has a been heavily overblown. and unlike golf course for value delivered AI datacenters are getting more not less efficient.
There is an implicit judgement. about which use is more acceptable which underpins this all
Yes, and my point is, a) “heavily overblown” is frequently, grossly inaccurate at a local level (where it matters), b) “more or less valuable” is only half the question, when “objectively harmful” is on the table. No more argument here. Just a plea to keep the analytical rigor when the stakes are as high as they are here. Cheers.
A question that came up in a recent conversation.
Is the “used” water by air data centers evaporated or flushed back into river/ lake/resevoir? And is it dirty?
People who live within 1 mile of a golf course are x times more likely to have cancer. So even though a portion of the water used on golf courses is run off, it’s heavily contaminated with chemicals.
How do data centers compare with that?
Older datafenters evaporate. newer ones will use closed loop
systems or direct systems. very new ones will use liquid immersion cooling.
That data about local temp increases is astonishing. Are we sure the study’s authors are correct? And how does it compare to other industries? A 9 degree celsius increase is catastrophic
Doesn’t pass a sniff test. And it badly fails the Andy Masley test: https://andymasley.substack.com/p/data-centers-heat-exhaust-is-not
Yep this makes sense
Exactly. Forget climate change modelling then and revert to mapping locations of data centres and tracking extreme weather patterns locally.. if only it was that easy
We have issued a correction, thanks everyone
Using 2023 #s for data center water cooling is substantially out of date, right?
Also, just because golf courses do it 30x more, it’s ok for data centers to guzzle 17B gallons of water, which presumably will double to 35B+ in short order, many times in areas where water table is already at the tipping point, eg Google DC in Dalles, Oregon ? Not convinced by this line of reasoning. At all.
pick you comparative: almond production, fracking, avocados, toilet leaks...
Yes, I appreciate that we already waste water elsewhere. But a) averages have little meaning here, and b) bad is still, uh, bad. ;) When it’s your small-town watershed and Google moves in to consume an extra 1 million gallons per month, quadrupling water use and depleting the water table 10x faster than otherwise (draining the last creek), talking about almonds doesn’t help solve the problem. Nor does using more of an overused resource make it less bad because someone else overused it first. But I’m all for taxing golf courses and AI inference to pay for watershed restoration ;)
That i think is the point. first of all water use has a been heavily overblown. and unlike golf course for value delivered AI datacenters are getting more not less efficient.
There is an implicit judgement. about which use is more acceptable which underpins this all
Yes, and my point is, a) “heavily overblown” is frequently, grossly inaccurate at a local level (where it matters), b) “more or less valuable” is only half the question, when “objectively harmful” is on the table. No more argument here. Just a plea to keep the analytical rigor when the stakes are as high as they are here. Cheers.
Energy + Water Tax -soon coming to a place near you.
🤔 short-term employment, long-term under-employment, the planet turns into a desert; who the hell is AI being built for?