keepthepace

When you make an urban park, you have to accept two things though. First, locally the city becomes more expensive. And second, it will occupy more space over potentially natural zones (over suburban area actually, that themselves will grow over peri-rural, that will push these, etc.)

If your goal is wildlife preservation or carbon capture, you probably want one more hectare in a forest rather than an hectare in a park inside a city.

Don’t get me wrong, I wanted more trees when I was living in a city. But the choice is between livable big cities or trying to make the cities as small impact as possible.

And also, water shortages have other causes, much more systemic, that I feel the datacenter scapegoat is a convenient distraction from.

You can cool chips with air, you can cool them with the sea or with non drinkable water. If they really get built in places where water is scarce, the problem is why the hell are they incentivized for that?

Water is plentiful and there should be no shortage of it. Where it is lacking it is either an environmental problem (desert areas should not be supposed to sustain cities) or a public infrastructure problem.

Seeing such a post upvoted on solarpunk gives me hope.

I don’t think the space race and space habitat is the way forward, but there is a future timeline that is sustainable and technologically advanced.

I am happy that some people still believe in it.

and save room for wild nature.

Arguably space habitat is the ultimate way to achieve that goal on earth :-)

Well, if you assume that the air recycling will be done by plants, you are going to have a need for a huge biosphere and probably for a lot of oceanic biome. If you plan on having an agriculture, if you plan on having pollinating insects, you are going to need a huge biomass.

You’re going to have far more nature than you will have humans.

This setting usually assumes that you have a fast transport network built in the core infrastructure. So being spread out is less of a problem, especially if you have free electricity through solar panel. That means that moving stuff around on electric motor is going to be basically free.

Absolutely, that’s why I’m mentioning it is a theoretical coefficient. It is worth noticing that the compressor, which is the part that’s being replaced here, is the main user of energy in a typical heat pump. So there is hope that if we get to the same level of engineering efficiency, we can get a very efficient system in the future. Maybe not up to 20, but certainly better than what we have now, and with a mechanically very troublesome part gone.

If you already know how typical heat pumps work, said video could be summed up by this illustration:

Interesting additional tidbit: theoretical coefficient of power is around 20 for this tech vs 3 for what we use. Problem: heat is absorbed and dissipated at the same place, so you still do need something that circulates it around, even if you don’t need a compressor anymore.

Just to know if you are not aware, you are putting a penny in a hot debate in the free software community on which license is the more open, is the best.

The MIT is clearly the most permissive because it allows you, among other things, to just run with the software and close it, adding your modification and sell it without sharing source.

Afero GPL prevents you from selling the software or even selling services that run the software without sharing/publishing it.

In a MIT->AGPL swap you will find people who consider it a step into being closer to free software ideals and people who consider it getting further away from it.

I think I already posted it there.

About regional communities, slightly tangential but in case there is somewhere a list of solarpunk communities hosted on different instances, solarpunk@jlai.lu is (tentatively) a french speaking solarpunk community

ASIC just means “specialized chip”. They don’t claim anything else.

You know what I want? A Whisper chip. Or whatever language is better now, but it is good enough for so many application. Give the sense of hearing to appliances. I guarantee you that in 20 years it will still be used.

I am not sure, I have tried to avoid this whole situation in the last few years :-) IIRC it can have its own CUDA version, but double check that.

The CUDA version is what matters the most (assuming you are on NVidia). Later CUDA versions have optimizations that earlier don’t, this may in turn dictate the actual driver version you can use.

I guess some models will simply deactivate some optimizations if you don’t have an appropriate version, though I mostly am aware of them failing in that case :-/

If you compare a model running on CUDA 11 vs a model running on CUDA 12, people may point out that it could be unfair, though this is generally nitpicky.

If you are worried about your perfs not being optimal, look in the log for messages like “<fast attention scheme XYZ> was deactivated because <cudaSuperOptimizedMegaSparseMatMult> was not available”

Dut to individual cars becoming the norm