I find this debate a bit odd. Climate change is a global issue. So moving these emissions out of Ireland so Ireland can reach it's targets doesn't help if those emissions mearly happen elsewhere in the world. In reality, there is two options either these data centers are simply not built, which isn't going to happen. Or rich countries such as Ireland, take it on to themselves to see taken on datacenters as a burden they must carry, coupled with increased renewables. With Ireland's wind on the west coast potential, this is something we must Especially with a temperant climate
Moving the datacenters away from East coast Dublin to West coast. Galway in an ideal world would be also useful.
>> moving these emissions out of Ireland so Ireland can reach it's targets doesn't help if those emissions mearly happen elsewhere
They aren't being moved out. This is basic market forces, negotiation between provider and consumer. Ireland would be perfectly happy to host these datacenters should they be carbon neutral, perhaps by constructing their own green energy support system. The datacenters don't want to do that. They want cheap electricity. They want to be simple electricity consumers and thereby outsource any emissions issues to the local grid. Ireland has chosen not to take on that burden. If another country is willing, then so be it. Nobody should engage in a race-to-the-bottom on emissions. If these datacenters want to be in Ireland, they are free to foot the bill.
I'd be ok with it. As long as nobody that's contracted to build housing or apartments here is involved. We'd somehow have meltdown before they laid the foundation.
I think others have already alluded to this, but the source of the electricity is the important bit. A datacenter running entirely on hydroelectricity is significantly greener than one running on coal. Moving the datacenter to a location where the source is as green as possible can have a big impact on emissions.
Exactly. Furthermore, the fine article mentions 21% going to datacenters (of which, which fraction might be off-shoreable anyway?), and 18% to homes.
Okay, that leaves 61% going to what?
Which other industry uses have been going up or down? Then compounded by "If we already had lots of wind and lots of solar, it wouldn’t be a problem", which, as far as datacenters go probably misses the need for storage.
What this is seems more of a meme article, linking presumably evil datacenters, with climate change, with "policy of low corporate taxation."
Commercial maybe? I used to work in a Lidl and it was mentioned when they started automating their store lights inside and outside to turn off at night that it was taking a large load of the grid.
Lot's offices, SME's etc?
Warehouses? Not sure if they count as industrial or not.
Of course. And metals smelters, and street lighting. Whatever.
The point was that the fine article didn't care about that because it was a political hatchet job, rather than informing on how or why electric power is used in Ireland.
Presumably you'd keep evicting them out of these tax havens until the only option they've got left is to innovate and optimize energy use, or heavily invest into renewable sources that don't have a huge heat/pollution footprint.
> either these data centers are simply not built, which isn't going to happen
And this is why we will never solve the climate issue, because we still refuse to even entertain the idea of doing what we must do, what we must have been doing for decades but have refused to so it's going to be harder and harder to quit.
Data centres consuming electricity isn't the problem. The problem is running those data centres on fossil fuels.
There has been an increased push to run the data centres on green technologies like solar, wind, and nuclear, e.g. [1]. Due to the machines being co-located makes it easier to do this compared to every website being hosted on machines spread over the country, where it is more likely that they will be run by fossil fuels.
Another problem is water consumption. People don't realize it, but these data centers also consume incredible amounts of fresh water.
There are a handful that run off of grey water, and it varies somewhat - but overall the consumption is enormous. It's a big problem for water-starved areas.
I wouldn't call it enormous. Some datacenters run dry cooling. A large datacenter (100MW) that uses evaporative cooling probably uses ~1000 acre-ft/yr, on par with a medium farm.
How do they consume the water exactly? If it’s just heat transfer, couldn’t it be reused once it passes through the data center? I would assume a small and cold country like Ireland could make good use of warm water, especially given its #2 ranking on tea consumption per capita!
A datacenter could consume a lot of water with evaporative cooling. I don't know how prevalent it is, but given how cheap and efficient evaporative cooling is, I'd guess datacenters use it a lot where possible (probably in combination with other cooling methods).
A datacenter should be condensing as little as possible. The cold side (where condensation occurs) is the inside, but you don't want your cooling system dripping on your racks.
Ireland is upping its game to take advantage of it being a very ‘windy’ country. A 130 tonne flywheel, spinning inside a vacuum at 3000 rpm is set to offer some inertia to stabilise the grid, a common problem with wind power production. We’ll need this if we’re to meet power demands and achieve the Irish government’s lofty wind power ambitions.
I miss the cultural era that was enthusiastic about economic growth. Most stories I see on HN, and other places, now react with reflexive condemnation of any kind of human industry. I worry about where this zeitgeist will lead us.
I'm confused as to why growth and progress as a species is tied inexorably to the mindlessly wasteful consumption of fossil fuels to the detriment of our own ability to live comfortably and feed ourselves. aren't we clever? can't we progress without screwing up the planet?
Right? That's the problem. There is nothing inexorable about it. France ages ago noticed a problematic reliance on fossile fuels and decided to actually solve that problem (for power and heating). That was a super ambitious program of building an entire, country-size industry capable of designing, building, fueling, running and reprocessing grid-scale nuclear reactors. It even worked. Growth WHILE solving problems. But yeah, in France as elsewhere, that effort was and is tied to political will and electoral moods. And so now it's in jeopardy.
Where did this cliche come from? I see it repeated everywhere from minds clearly not creative enough for said notion.
Wishful thinking that they can just blame deep structural issues like oil dependency and problematic chains of logistics on rich people and that in their absence all problems would magically solve themselves and flowers would bloom?
I think this is a very serious problem. There are still ambitious people, but ambition is now ipso facto evil (whether in business, tech, science, finance, law). That's the politically correct meme. Automatic. And hard to reverse as nieces and nephew have reported constant "save the planet" themes in school work. With "shrink the economy" being the only solution ever suggested.
When whichever cause may be yours (climate, housing, faster cat photos) might actually benefit from ambition, scale, technology focused progress, better ways to even think about anything and everything.
The problem is not growing the economy per se. The problem is the economy has always grown by making profit of externalising negative impacts. Be it environmental, humanistic.
We have to learn to grow but to do it fair. Fair for the environment and for the people (meaning not having the profits end up with the 0.01% while the workers toil for minimum wage). Only then can we really grow as humanity.
The problem is what we call "ambition" is just greed. And that keeps us in this circle of destruction.
Ah, yes that mythical city on the hill "fairness" which is a clear and static standard and not at all a rapidly moving goalpost. Along with its equally mythical counterpart of "the people" whose viewpoints are not a convenient reflection of whatever position is most convenient to the speaker. That is truly the way to engage in the hardest of tasks already prone to failure, chasing fucking unicorns!
This, horribly, may be a current Silicon Valley view. That's the problem.
By contrast the example I gave elsewhere in this discussion is "France ages ago noticed a problematic reliance on fossil fuels and decided to actually solve that problem (for power and heating). That was a super ambitious program of building an entire, country-size industry capable of designing, building, fueling, running and reprocessing grid-scale nuclear reactors. It even worked. Growth WHILE solving problems. But yeah, in France as elsewhere, that effort was and is tied to political will and electoral moods. And so now it's in jeopardy."
That was a giant program with participation of many private companies - not even one private one dominating costs or profits that I noticed. The newly created companies forming the project started for some, and ended up for others owned by France through many different channels: national research labs, national electric utility, development banks, government agencies, etc, etc. No doubt some private companies made excellent money from that project - but that is not its most prominent feature by far.
It was an ambitious engineering and organizational effort.
I miss that era too. But the shift isn’t purely cultural: it’s an acknowledgement of the underlying reality that climate change is getting to be real bad. I worry about where a growth-at-all-costs mindset has lead us at least as much.
We were causing mass extinctions with pointy sticks as the state of the art, in charismatic mega fauna no less. As common the mythical era of "not causing mass extinctions" never even existed.
Even when humanity was at its lowest creating the genetic bottleneck we were nomadic because we ate the prior area empty. Now we've progressed to sedentary lifestyles being possible long term and even soil management. The only path out is up.
> We were causing mass extinctions with pointy sticks as the state of the art
Not at anything like this scale, nor this rate, as is clear from the wiki article linked above.
> the mythical era of "not causing mass extinctions" never even existed.
It existed everywhere we weren't, and again, scale and rate both matter.
> Even when humanity was at its lowest creating the genetic bottleneck we were nomadic because we ate the prior area empty.
The current consensus is a lot more complex than humans being "purely nomadic" because they "ate the prior area empty."
Evidence of early permanent settlement appears much earlier than the advent of agriculture ~12,000 years ago. Ohalo II in Israel is 23,000 years old. There's likely many other sites as old and older, but our archaeology is not advanced enough to know more.
> Now we've progressed to sedentary lifestyles being possible long term and even soil management
... What?? We're only now rediscovering soil management practices which were used hundreds/thousands of years ago by people native to areas we've all but destroyed. If we'd listened to them instead of dehumanizing and murdering them things could be different, but we can still at least not whitewash them out of history.
Globally speaking, top soil is in a variety of concurrent crises. It's in the worst shape it's ever been.
Reflexive? Take a look in the mirror. The condemnation is based on arguments that you can try to refute; or you can be reflexively enthusiastic, it’s up to you.
Eh, there's a lot of enthusiasm for all those battery threads.
I think there's really two factors going on.
1) Everything has trade-offs and a lot of the economic growth comes with a trade-off they don't like. The top comment in this thread isn't even "screw economics", it's "I wish they built them where the wind power was". What does the N+1 data center even get us as a society while a rocket to the moon is literal science fiction.
2) Some of the "economic growth" is really just stagnation at a cheaper price. Firing ~10% of your workforce because AI-tooling (allegedly) lets you produce the same amount with less people isn't growth.
I'm old enough to remember the world without a constant presence of the internet and connectivity. I wonder how much energy these datacenters have actually saved thanks to technologies like google maps, email, online shopping, etc..
I still get a visit from the mailman in his little truck six days a week. So email has saved nothing there.
I get a lot more other deliveries too, but maybe I'm driving less to offset that? Haven't tracked this closely enough to say.
Google Maps probably has saved some fuel in that I take fewer wrong turns and don't have to double back quite so often when I'm in an unfamiliar area. Again I just have my own impressions, but GPS tech in general probably is a net win in terms of energy saved.
Email saving nothing isn't true at all. The victorians had multiple times a day mail service and newspapers. The mail services previously bemoaned a loss of letters before retail got displaced by online shopping. Remember the concept of the corporate Mail Room? A hungry place that employment could always be found in? The volumes of mail involved were vastly more in the past.
The consumption now and in near future will be mostly driven by large AI models, to generate some crazy images, scam some people, and maybe also help individuals and companies a bit. I am not holding my breath on some overall positive effect.
In past few years, it has been crypto mining ( maybe not in IR but globally definitely yes), I think we can all agree a pointless emotional fomo endeavor.
That online shopping is also debatable, people often order low quality shit they don't really need because its ridiculously cheap. Manufacturing, transport, packaging, not so great for the planet.
The days before the internet, I recall them too. Sure, you had to get to the shop, but I had way less clothing and overall everything and I kept it for longer (I am still very much an un-fashionate outlier of modern era but even for me its headed in bad direction).
Quebec does have a few. Amongst the issues is that scaling hydro power is hugely expensive, is riddled with red tape, and over all takes too long. Meanwhile you could build a datacenter and a solar farm in Texas and be up in 2 years.
Governments ought to incentivize efficient datacenter workloads, the same way they're incentivizing heat pumps over resistive heat. If they can give people tax credits for having efficient homes, they can give maintainers of large-scale software something in exchange for optimizing that software. I bet some very large percentage of data center electricity usage is directly attributable to absolute bloat, like VMs that could be containers, build pipelines that chug away endlessly producing nearly identical build artifacts every time, dumb polling that could utilize a smarter message queue push, neural nets that could be if statements, etc. -- where is the "LEED Gold" of software?
There may ultimatley need to be taxes on wasteful energy use. Less so on infastructure but very high on frivolous energy waste for things like crypto and AI as absurdly wasteful toys without meaningful use.
Moving the datacenters away from East coast Dublin to West coast. Galway in an ideal world would be also useful.