CUDA

2026-05-30 Posts

Enabling P2P Communication on NVIDIA RTX Servers: From Driver Patching to Performance Verification

A detailed guide on how to enable Peer-to-Peer (P2P) communication for RTX GPUs by modifying NVIDIA kernel modules on Debian/Ubuntu systems, and verifying bandwidth using CUDA Samples.

Technical Tutorial GPU

#NVIDIA #RTX #P2P

2026-05-19 Posts

Compiling llama.cpp on Linux: Full Guide from CPU to CUDA Acceleration

A detailed guide on how to compile llama.cpp from source on Linux, covering basic CPU versions and NVIDIA GPU (CUDA) acceleration configuration steps. Includes complete compilation command reference.

#llama.cpp #Linux #Compilation Guide