Worked at a startup that unfortunately went under, trying to reuse heat generated in datacenters. (saves on electricity spend on airco and saves on having to generate heat elsewhere)
The temperature differential between what the hottest chips can accept, about 70C, and the environment suggests this is a highly inefficient source of energy, due to Carnot limits.
I wish people would go for the big guns, fully renewable energy, nuclear, electric transport, green concrete and steel - the major issues instead of wasting time and money recovering 10% of what some coal power plant supplied to their data centre.
It's better than nothing but very close to nothing. It's a pervasive way of thinking in green engineering, every bit helps but what we need is BIG help:
https://www.withouthotair.com/c19/page_114.shtml
I wish people would go for the big guns, fully renewable energy, nuclear, electric transport, green concrete and steel - the major issues instead of wasting time and money recovering 10% of what some coal power plant supplied to their data centre.
It's better than nothing but very close to nothing. It's a pervasive way of thinking in green engineering, every bit helps but what we need is BIG help: https://www.withouthotair.com/c19/page_114.shtml