One year ago, I predicted in Videos and articles that Elon Musk would build AI data centers on Mars. A few days ago, Elon Musk confirmed the general plan for AI data centers on Mars to an All hands meeting of XAI.
I also wrote one third of the book AI Rising. The book is available on Amazon and is a best seller in the AI category.
I wrote a chapter on AI data centers on Mars to enable about 3 trillion AI chips.
The book goes over the important future aspects of AI in detail. The chapter that some might have considered to be the craziest is now confirmed as part of the Elon Musk plan.
Mars has a ~24.6-hour day, so solar power isn’t constant without storage. Assuming sufficient energy storage (e.g., batteries), the average power over time matches the intercepted power from the sunlit side, keeping our calculation valid.
Dust Storms: These can reduce solar flux, but for a theoretical maximum, we use the average flux of 586 W/m².
Orbital Variation: Mars’ solar flux varies from 492 W/m² to 717 W/m² due to its elliptical orbit, but 586 W/m² is a reasonable average.
Transmission Losses: A planet-wide grid would incur losses, but we assume the harnessed power is delivered efficiently, with overhead included in the PUE (Power Utiliziation Efficiency).
With 20% efficient solar panels covering the sunlit side of Mars, Elon Musk, XAI, SpaceX and Tesla can harness 4,240 terawatts of power. Accounting for a PUE of 1.2, this supports approximately 3.53 × 10¹² (3.53 trillion) 1000-watt GPUs.
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
Datacenter in space is nonsense. datacenter on Mars is so absurd that makes the development of the cybertruck a wise business choice.
Are these AI centers only for training?
Users will not like LLMs that take 10 minutes (due to speed of light) to answer a question
you predicted tesla’s share price would be 1000 – 1500 in 2025
I said that could happen IF unsupervised robotaxi was deployed and there was a surge in car demand from that. I had videos indicating that this is deferred for 6 months based upon those delays
Plans need to be based on reality. This sounds like his FSD. Ask an AI to list all predictions made on it since around 2015. It seems to be stuck around the 85 percent done for years now. That is a real sign that they do not yet have a real path to success.
Seems like an opportunity for full scale Solar Power satellites beaming power via microwave or laser to the surface from areosynchronous orbit.
Not much atmosphere to get in the way. No birds, planes, neighbors to worry about. Vast demand justifying scaling. Build in Earth orbit and send to Mars.
Indeed, SPS on Mars is easier than Earth; Synchronous orbit is much lower. And both Phobos and Deimos are there to supply building material with only 1-200 mps delta V.