Tag archive
Banana Pi's BPI-AI2N Packs: Banana Pi's BPI-AI2N runs vision models on-device: a Renesas RZ/V2N module with 15 TOPS, 8 GB of RAM, dual GbE and two MIPI CSI camera inputs.
As part of my 2026 learning goals around Java on RISC-V (see this post about x86 versus ARM versus...
How we discovered and fixed the root cause preventing Next.js from recognizing riscv64 architecture, achieving the first successful native SWC build o
Running LLM inference on RISC-V with the Python ML stack. Transformers generates text at 0.15-0.93 tok/s on a BananaPi F3, while llama.cpp hits 8.21 t
How I built a PEP 503 wheel index for riscv64 Python packages using two BananaPi F3 boards, GitHub Actions self-hosted runners, and 26 forked repos.
A clean-machine pip install exposed stale riscv64 PyTorch wheels, missing system libraries like libopenblas, and transitive dependencies (numpy, grpci
Building the entire LLVM/Clang WASI SDK toolchain from source on a RISC-V single-board computer for self-hosted WebAssembly development.
Building llama.cpp from source on a RISC-V board and running a local LLM inference server with TinyLlama 1.1B, achieving 8.5 tokens/second.
Benchmarking llama.cpp on SpacemiT's K3 RISC-V chip with three models: X100 cores hit 76 t/s, A100 AI cores are 30x slower on pp until vendor IME2 cod
How we packaged Docker Engineering's cagent multi-agent AI runtime for RISC-V64, creating the first Linux distribution package for this powerful tool
Spoiler first: the Banana Pi M7 is not actually $80. Street price in 2026 is $165 on AliExpress, $210...
Banana Pi — Como Habilitar o SSH no Raspbian Henrique...