Comment on ChatGPT 5 power consumption could be as much as eight times higher than GPT 4 — research institute estimates medium-sized GPT-5 response can consume up to 40 watt-hours of electricity

A_norny_mousse@feddit.org ⁨7⁩ ⁨months⁩ ago

• ThePinkUnicorn@lemdro.id ⁨7⁩ ⁨months⁩ ago

  For training yes, but during operation by this studies measure Deepseek actually has a higher power draw, according to the article. Even models with more efficient programming use insane amounts of electricity

  This was higher than all other tested models, except for OpenAI’s o3 (25.35 Wh) and Deepseek’s R1 (20.90 Wh).

  source

  • A_norny_mousse@feddit.org ⁨7⁩ ⁨months⁩ ago

    OK I guess I didn’t read far enough but your quote says that Deepseek uses less than Open AI?

    source

    • ThePinkUnicorn@lemdro.id ⁨7⁩ ⁨months⁩ ago

      Less than Open AI’s o3, but that’s because o3 was estimated to use even more power than GPT 5’s 18 Wh per query.

      source
• joonazan@discuss.tchncs.de ⁨7⁩ ⁨months⁩ ago

  My guess would be that using a desktop computer to make the queries and read the results consumes more power than the LLM, at least in the case of quickly answering models.

  The expensive part is training a model but usage is most likely not sold at a loss, so it can’t use an unreasonable amount of energy.

  Instead of this ridiculous energy argument, we should focus on the fact that AI (and other products that money is thrown at) aren’t actually that useful but companies control the narrative. AI is particularly successful here with every CEO wanting in on it and people afraid it is so good it will end the world.

  source
• kautau@lemmy.world ⁨7⁩ ⁨months⁩ ago

  And water which will also increase as fires increase and people have trouble getting access to clean water

  techhq.com/…/ai-water-footprint-suggests-that-lar…

  source

  • FauxLiving@lemmy.world ⁨7⁩ ⁨months⁩ ago

    It would only take one regulation to fix that:

    Datacenters that use liquid cooling must use closed loop systems.

    The reason they dont, and why they setup in the desert, is because water is incredibly cheap and energy to cool a closed loop system is expensive. So they use evaporative open loop systems.

    source

    • Jason2357@lemmy.ca ⁨7⁩ ⁨months⁩ ago

      Closed loop systems require a large heat sync, like a cold water lake, limiting them to locations that are not as tax advantageous as dry red states.

      source

      • NikkiDimes@lemmy.world ⁨7⁩ ⁨months⁩ ago

        Aw, that’s unfortunate for the big mega tech corps. Anyway.

        source
    • Ilovethebomb@sh.itjust.works ⁨7⁩ ⁨months⁩ ago

      That increases your energy use though, because evaporative cooling is very energy efficient.

      source

      • FauxLiving@lemmy.world ⁨7⁩ ⁨months⁩ ago

        We can make energy from renewable sources.

        Fresh drinking water is finite, especially in the desert.

        source
    • kautau@lemmy.world ⁨7⁩ ⁨months⁩ ago

      Unfortunately I wonder if it’s more expensive to set up a closed loop system that’s really expensive or to buy lawmakers that will vote against bills saying you should do so and it’s a tale old as time

      source

      • FauxLiving@lemmy.world ⁨7⁩ ⁨months⁩ ago

        Politicians are cheap

        source

        • -> View More Comments
• rdri@lemmy.world ⁨7⁩ ⁨months⁩ ago

  Also don’t forget how people like wasting resources by asking questions like “what’s the weather today”.

  source
• ThePowerOfGeek@lemmy.world ⁨7⁩ ⁨months⁩ ago

  BTW a lot of it seems to be just inefficient coding as Deepseek has shown.

  Kind of? Inefficient coding is definitely a part of it. But a large part is also just the iterative nature of how these algorithms operate. We might be able to improve that via code optimization a little bit. But without radically changing how these engines operates it won’t make a big difference.

  The scope of the data being used and trained on is probably a bigger issue. Which is why there’s been a push by some to move from LLMs to SLMs. We don’t need the model to be cluttered with information on geology, ancient history, cooking, software development, sports trivia, etc if it’s only going to be used for looking up stuff on music and musicians.

  But either way, there’s a big ‘diminishing returns’ factor to this right now that isn’t being appreciated. Typical human nature: give me that tiny boost in performance regardless of the cost, because I don’t have to deal with. It’s the same short-sighted shit that got us into this looming environmental crisis.

  source

  • kescusay@lemmy.world ⁨7⁩ ⁨months⁩ ago

    Coordinated SLM governors that can redirect queries to the appropriate SLM seems like a good solution.

    source

    • sleep_deprived@lemmy.dbzer0.com ⁨7⁩ ⁨months⁩ ago

      That basically just sounds like Mixture of Experts

      source

      • kautau@lemmy.world ⁨7⁩ ⁨months⁩ ago

        Basically, but with MCP and SLMs interacting rather than a singular model, with the coordinator model only doing to work to figure out who to field the question to, and then continuously provide context to other SLMs in the case of more complex queries

        source
    • JoeKrogan@lemmy.world ⁨7⁩ ⁨months⁩ ago

      Powered by GNU Hurd

      source