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

    AI chips shipped through 2027 could run tens to hundreds of millions of concurrent frontier-model agents

    A quoted blog passage estimates that nonstop agents could supply as many weekly working hours as about 140–720 million full-time employees.

    T(
    AC
    IR
    6 Sources, ,

    TLDR

    A blog excerpt shared by a user projects tens to hundreds of millions of concurrent frontier-model agents running on AI chips shipped through 2027, supplying as many weekly working hours as about 140–720 million full-time employees. It says more efficient models could potentially support billions on the same hardware. Applying DeepSeek V4 Pro serving benchmarks to projected hardware supply yields an estimate of approximately 1.9 billion concurrent agents, supplying as many weekly working hours as 8 billion people working 40 hours.

    Combined views

    36.9K

    6 Sources, first seen 4h ago

    Combined views

    36.9K

    6 Sources, first seen 4h ago

    693 likes
    Chart: potential concurrent agents from 2025–27 memory shipments by model, from 30–60 million (Claude Fable 5) and 97–195 million (GPT-5.6 Sol) to 240 million (Kimi K3), 571 million (GLM-5.2) and 1.9 billion (DeepSeek V4 Pro).
    4h ago
    first seen 4h ago
    693 likes
    40 comments
    196 saves
    135 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Featured Source
    40 comments
    196 saves
    135 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    #7

    Today's Rank

    #7

    6 Sources

    @AndrewCurran_From the blog: 'AI chips shipped through 2027 could run tens to hundreds of millions of concurrent frontier-model agents. Running nonstop, these agents would supply as many weekly working hours as about 140–720 million full-time employees. More efficient models could potentially support billions of agents on the same hardware. Applying DeepSeek V4 Pro serving benchmarks to the projected hardware supply yields approximately 1.9 billion concurrent agents supplying as many weekly working hours as 8 billion people each working 40 hours.4h
    @teortaxesTexto be clear, even V4 Pro is slow af This means that with V4.1 Flash or similar architectures, you could run the entire human population's worth of agents on this capacity. I think even DeepSeek's own DSec by eoy 2027 will be doing tens of millions of concurrent agents.4h
    @zephyr_z9@teortaxesTex that's the plan4h
    @irinarishRT @EpochAIResearch: We estimate AI infrastructure could soon support hundreds of millions or even billions of AI agents. At the high end,…3h
    @HaydnBelfieldEpoch back at it again with some of the best and most interesting analysis in the world3h
    @ChrisPainterYupEquivalent full-time digital remote human workers (in this case 140-720 million) is going to be a good way to think about AI for the next little stretch.2h

    6 Sources

    @AndrewCurran_From the blog: 'AI chips shipped through 2027 could run tens to hundreds of millions of concurrent frontier-model agents. Running nonstop, these agents would supply as many weekly working hours as about 140–720 million full-time employees. More efficient models could potentially support billions of agents on the same hardware. Applying DeepSeek V4 Pro serving benchmarks to the projected hardware supply yields approximately 1.9 billion concurrent agents supplying as many weekly working hours as 8 billion people each working 40 hours.4h
    @teortaxesTexto be clear, even V4 Pro is slow af This means that with V4.1 Flash or similar architectures, you could run the entire human population's worth of agents on this capacity. I think even DeepSeek's own DSec by eoy 2027 will be doing tens of millions of concurrent agents.4h
    @zephyr_z9@teortaxesTex that's the plan4h
    @irinarishRT @EpochAIResearch: We estimate AI infrastructure could soon support hundreds of millions or even billions of AI agents. At the high end,…3h
    @HaydnBelfieldEpoch back at it again with some of the best and most interesting analysis in the world3h
    @ChrisPainterYupEquivalent full-time digital remote human workers (in this case 140-720 million) is going to be a good way to think about AI for the next little stretch.2h