AnalysisAI ModelsNovember 1, 2026

llama2.c runs on a 2006 iPod at 80 MHz with no FPU

Read original source →reddit.com

A Rockbox plugin runs Karpathy's llama2.c TinyStories models on an iPod Video 5.5G (PortalPlayer PP5022, ARM7TDMI at 80 MHz, 64 MB RAM). With no FPU, the int8 runq.c was ported to integer-only math using lookup tables for exp/softmax plus an integer square root.

1 source

More stories today

Open the live feed