If the subsidies were removed from fossil fuels and applied to solar it would transform the US energy landscape.

However this is a difficult proposition as solar is inherently a grassroots energy source that can be deployed by people at their homes, it's democratizing and distributes power, capital and control away from central power structures.

The fossil fuel industrial complex is the opposite.

Surely most deployed solar isn't owned by average people? I would expect that it's almost all coming from large-scale deployments, and a good investment opportunity.

Both comments are true. We could go faster if fossil subsidies shifted to solar and batteries, but we will still go fast regardless. Most US solar is utility scale, but buying your own solar is cheap enough now you can almost go off the grid (battery price decline will catch up shortly) assuming you have enough space for panels. Utility scale solar is still a good investment, even with the loss of tax subsidies, and is the fastest way to deploy new generation capacity.

Regardless, we’ve reached a global tipping point where solar, battery, and EV deployment continues to accelerate and peak fossil fuel demand is very near.

EU taxes fossil fuels heavily and the didn't really build more solar than the US.

What's the source of such plainly wrong "USA #1" posts? https://www.bruegel.org/analysis/transatlantic-clean-investm...

The generation of electricity from solar is more or less the same.

But they did build more, the USA is just further south.

Solar only works during the day. Datacenters need energy 24/7. Consumers who didn’t install solar with batteries will end up with higher prices.

At utility scale you can have pumped hydro and other economic battery alternatives.

CATL's announcements of Sodium Ion batteries promise an order of magnitude cost reduction over lithium. Having enough storage to capture peak energy will be game changing; the only significant hurdle is to modernize the grid but there's a lot of resistance to that /rimshot

Why should data centers need more than the usual proportion of energy at night?

We could also choose to just build solar powered datacenters and have a follow-the-sun model for the data processing. That small delay would still meet most of the needs of users. But nope, BigCorp isn't smart enough to be that innovative and instead doubles down on the old method and demands more nuclear power.

Electricity is cheap compared to GPU capital costs and depreciation. Intermittent usage is only cost efficient with capital that cheaply buffers energy either generically like batteries or specifically like heating up your smelters. It is not unsolvable, but not as easy as you imagine to retool.

Well they'll end up with higher prices at night. I wonder if we'll see a shift to energy-intensive processes being run during the daytime.