3 ms·

solar farms are worst kind of power source for constant loads like datacenters running AI training

spacington 4mo ago
Solar power still can take a certain amount of load from any other source and saving money and CO2 while doing so. And grid battery works and is cheap enough now.
actionfromafar 4mo ago
That's why we are getting clean, beautiful coal the likes of which nobodys ever seen before!
alex_duf 4mo ago
I'm hoping the kind of gas turbines being installed on these datacenters are capable of quickly responding to a load change, meaning the primary source of energy could be solar during the day and complement when there isn't enough energy. But I haven't looked into where these datacenter are being placed, I'm assuming that although solar is cheap now, the surface needed would make the purchase of nearby land probably not worth it. These new categories of datacenters are becoming very energy dense...
trhway 4mo ago
you're just incorrect. You probably missed 2 points : - battery storage - and in the article "However, AI labs and some hyperscalers have relaxed those requirements as there is now a lower uptime tolerance applied to both inference and training, not just training. Many of Meta’s self-built AI datacenters, for example, target just two nines of uptime and forgo backup generators entirely, as detailed in our Industrials Model."
numpad0 4mo ago
but you can't trust especially hyperscalers with securely sealed HEUs in shipping containers
adornKey 4mo ago
Why? Unlike loads involving a real physical process there is absolutely no need for AI-training to be constant.
- pu_pe 4mo ago
  If you invest into chips that deprecate in value really fast, not utilizing them to their full capacity because of power constraints would be counter productive.
- everfrustrated 4mo ago
  You are correct in the sense that they can stop work in a way many generic server use cases can't (which is seen in lowering power supply reliability requirements as the article mentions), but running expensive servers at 50% utilization would dramatically affect the revenue generated per capital invested - IE you couldn't afford to buy the servers.
  - ZeroGravitas 4mo ago
    If it just costs more, similar to how it's possible to have a low water use datacenter, then in a perfect spherical cow economic universe then the lower cost option is better. In the real world you'd need to add some cost to account for the externalities on water and GHG. Do the numbers still work out for the gas powered plant? If not then you're just exploiting unaccounted for externalities.

Solar power still can take a certain amount of load from any other source and saving money and CO2 while doing so. And grid battery works and is cheap enough now.

actionfromafar 4mo ago

That's why we are getting clean, beautiful coal the likes of which nobodys ever seen before!

alex_duf 4mo ago

I'm hoping the kind of gas turbines being installed on these datacenters are capable of quickly responding to a load change, meaning the primary source of energy could be solar during the day and complement when there isn't enough energy. But I haven't looked into where these datacenter are being placed, I'm assuming that although solar is cheap now, the surface needed would make the purchase of nearby land probably not worth it. These new categories of datacenters are becoming very energy dense...

trhway 4mo ago

you're just incorrect. You probably missed 2 points : - battery storage - and in the article "However, AI labs and some hyperscalers have relaxed those requirements as there is now a lower uptime tolerance applied to both inference and training, not just training. Many of Meta’s self-built AI datacenters, for example, target just two nines of uptime and forgo backup generators entirely, as detailed in our Industrials Model."

numpad0 4mo ago

but you can't trust especially hyperscalers with securely sealed HEUs in shipping containers

adornKey 4mo ago

Why? Unlike loads involving a real physical process there is absolutely no need for AI-training to be constant.

pu_pe 4mo ago
If you invest into chips that deprecate in value really fast, not utilizing them to their full capacity because of power constraints would be counter productive.
everfrustrated 4mo ago
You are correct in the sense that they can stop work in a way many generic server use cases can't (which is seen in lowering power supply reliability requirements as the article mentions), but running expensive servers at 50% utilization would dramatically affect the revenue generated per capital invested - IE you couldn't afford to buy the servers.
- ZeroGravitas 4mo ago
  If it just costs more, similar to how it's possible to have a low water use datacenter, then in a perfect spherical cow economic universe then the lower cost option is better. In the real world you'd need to add some cost to account for the externalities on water and GHG. Do the numbers still work out for the gas powered plant? If not then you're just exploiting unaccounted for externalities.