GPU LLM Inference Sharding
A homelab-scale experiment in running larger local language models by sharding inference across affordable used GPUs, PCIe layouts, and networked nodes.
All the various projects, new and old, I've worked on.
A homelab-scale experiment in running larger local language models by sharding inference across affordable used GPUs, PCIe layouts, and networked nodes.
An ongoing experiment in learning machine learning through geometry processing, retopology, and artist-friendly Python tooling.
An exploration of full pathtraced raytracing with various techinques
A short material study focused on building convincing bark detail and surface breakup in Substance Designer.
A systems-oriented prototype showing how the command pattern can drive animations, sequences, and reusable gameplay-style behaviors.
A full pipline exploration from procedural Houdini geometry to pathtraced render.
A basic 2D procedural level generator in Unity focused on repeatable rules, layout variation, and practical iteration.
An OpenGL rendering study centered on texture pipelines, lighting passes, and assembling a more complex final frame.
A foundational OpenGL rendering exercise loading OBJ models with Phong lighting and a simple three-point light setup.
A lightweight particle system in a custom 2D engine exploring compute-driven updates, simple vector motion, and orbit behaviors.