A 10-node Raspberry Pi 5 cluster built with 16GB CM5 Lite modules topped out at 325 Gflops - then got lapped by an $8K x86 Framework PC cluster running 4x faster. On the bright side? The Pi setup edged out in energy efficiency when pushed to thermal limits.
It came with 160 GB total RAM, but that didn’t help much. AI workloads like LLaMA 70B dragged 5–25x slower than on x86 builds. Blame the weak iGPU, blame the interconnect bottlenecks - they both kill throughput.
