Running CPUs or GPUs in orbit is technically feasible. Scaling, transport, and servicing are the harder problems.
Running CPUs or GPUs in orbit is technically feasible. Scaling, transport, and servicing are the harder problems.
More from this video
See all →The human brain runs on roughly 5–20 watts, making it about a million times more energy-efficient…
U.S. data centers use roughly as much water annually as California almond farms use in four days.
Traditional racks used about 10 kilowatts. AI racks can consume roughly 100 kilowatts.
Fully closed-loop systems using car-like liquid coolants can consume nearly zero water from outside.
In 2016, Google used AI to study data-center power and cooling, improving efficiency by 40%.
U.S. data centers consumed about 17 billion gallons of water annually a couple of years ago.
Radiation, power, cooling, and mission duration are considered solvable. Launch economics remains…
New AI data centers increasingly use closed-loop cooling, recirculating water or coolant instead…
Don't lose this one
A free account saves insights like this to your Boards, and Korva resurfaces them so you actually remember.