• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    A user estimates simulating a human brain would take roughly 100 megawatts

    The user puts a human with an IQ of 100 at 20 watts and says simulating a human brain would require a large data center.

    RO
    BP
    PD
    7 Sources, ,

    TLDR

    Asked “How did you estimate this?”, a user replied with a power comparison: 20 watts for a human with an IQ of 100, versus a large data center using roughly 100 megawatts to simulate a human brain.

    Combined views

    159.8K

    7 Sources, first seen 17d ago

    870 likes

    Combined views

    159.8K

    7 Sources, first seen 17d ago

    870 likes
    17d ago
    first seen 17d ago
    79 comments
    213 saves
    67 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    79 comments
    213 saves
    67 reposts
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    7 Sources

    @pmddomingos@AravSrinivas Human: IQ = 100, power = 20 W Power required to simulate a human brain: large data center, ~100 MW
    @BorisMPowerThis number is way off. You’re comparing one human brain to an entire data center capable of training new models. The actual answer is: IQ points per watt: Humans: 5 AI: 7 - 40 (!!) By my rough calculations the current AI is already served more efficiently than humans for equivalently intelligent work performed! A single gpu can produce outputs and parse inputs way faster than a human and can do many of such requests in parallel. A larger model generally mostly means you need more gpus to serve it, but that often also increases how many requests you can serve in parallel, so I’m reporting on the share of the serving system, and also normalizing for speed. Really a better way to measure and report it is joules per equally good completed task, as your formulation suggests we can likely add more watts to get more iq, which isn’t correct. I did some back of the napkin calculations based on the latest serving hardware and the most powerful open source models. But even given pretty generous assumptions it turns outout computers come ahead. For example the request is allocated approximately 60 W while running, but finishes a task in about one-seventh of the human time.
    @timshi_ai@pmddomingos not sure i follow this argument: DeepSeek 4.1 Flash on B300: 91,833 total tokens/sec ÷ ~1,900W per GPU* ≈ 48 tokens/sec/watt. Brain: 5 tokens/sec ÷ 20W ≈ 0.25 tokens/sec/watt AI is far more efficient. https://inferencex.semianalysis.com/run/deepseek-v41-flash-on-b300
    @tszzlRT @BorisMPower: This number is way off. You’re comparing one human brain to an entire data center capable of training new models. The actu…
    @mayferpretty proud of humans for how we master a topic in 200 kWh

    7 Sources

    @pmddomingos@AravSrinivas Human: IQ = 100, power = 20 W Power required to simulate a human brain: large data center, ~100 MW
    @BorisMPowerThis number is way off. You’re comparing one human brain to an entire data center capable of training new models. The actual answer is: IQ points per watt: Humans: 5 AI: 7 - 40 (!!) By my rough calculations the current AI is already served more efficiently than humans for equivalently intelligent work performed! A single gpu can produce outputs and parse inputs way faster than a human and can do many of such requests in parallel. A larger model generally mostly means you need more gpus to serve it, but that often also increases how many requests you can serve in parallel, so I’m reporting on the share of the serving system, and also normalizing for speed. Really a better way to measure and report it is joules per equally good completed task, as your formulation suggests we can likely add more watts to get more iq, which isn’t correct. I did some back of the napkin calculations based on the latest serving hardware and the most powerful open source models. But even given pretty generous assumptions it turns outout computers come ahead. For example the request is allocated approximately 60 W while running, but finishes a task in about one-seventh of the human time.
    @timshi_ai@pmddomingos not sure i follow this argument: DeepSeek 4.1 Flash on B300: 91,833 total tokens/sec ÷ ~1,900W per GPU* ≈ 48 tokens/sec/watt. Brain: 5 tokens/sec ÷ 20W ≈ 0.25 tokens/sec/watt AI is far more efficient. https://inferencex.semianalysis.com/run/deepseek-v41-flash-on-b300
    @tszzlRT @BorisMPower: This number is way off. You’re comparing one human brain to an entire data center capable of training new models. The actu…
    @mayferpretty proud of humans for how we master a topic in 200 kWh