iFANN
    Rechercher sur iFANN...
    Connexion
    Accueil
    Actualités
    Vidéos
    Photos
    GIF
    Explorer
    Sondages
    Récompenses
    iFAMOUS
    Wiki
    Animé
    Salons
    Notifications
    Messages
    Favoris
    Profil
    WikiRécompensesiFAMOUSClassementsSecteursRécompenses créateursRécompenses utilisateursConditionsConfidentialitéRègles de la communautéRetrait / DMCAAideDéveloppeurs

    © 2026 iFANN

    Accueil
    Rechercher
    Messages
    Alertes
    Profil

    Publication

    Nate
    Nate@nate_512
    📱Kimi K3📱GitHub💭AI

    Kimi K3 runs on 8GB RAM

    WAIT... YOU CAN NOW RUN A TWO POINT SEVEN TRILLION PARAMETER MODEL ON AN 8GB LAPTOP 🤯 The Kimi K3 engine achieves this with a 176KB binary written in portable C. It bypasses memory limits by streaming the 1.56TB weights directly from disk for every single token. → 8GB RAM gets you 26 seconds per token → 128GB RAM gets you 5 seconds per token Same exact math and byte identical results regardless of the machine. Zero GPUs required. This is pure brutalist engineering. Free and open-source. repo in 🧵↓

    8h

    6 J'aime0 Je n'aime pas1 Republications1 Commentaires
    ?

    Commentaires

    Aucun commentaire pour le moment. Soyez le premier !

    Publication

    Nate
    Nate@nate_512
    📱Kimi K3📱GitHub💭AI

    Kimi K3 runs on 8GB RAM

    WAIT... YOU CAN NOW RUN A TWO POINT SEVEN TRILLION PARAMETER MODEL ON AN 8GB LAPTOP 🤯 The Kimi K3 engine achieves this with a 176KB binary written in portable C. It bypasses memory limits by streaming the 1.56TB weights directly from disk for every single token. → 8GB RAM gets you 26 seconds per token → 128GB RAM gets you 5 seconds per token Same exact math and byte identical results regardless of the machine. Zero GPUs required. This is pure brutalist engineering. Free and open-source. repo in 🧵↓

    8h

    6 J'aime0 Je n'aime pas1 Republications1 Commentaires
    ?

    Commentaires

    Aucun commentaire pour le moment. Soyez le premier !