How-to guides and AMD/ROCm optimization recipes for open AI-for-science models.

🚀 Automated installation script for vLLM on HPC systems with ROCm support, optimized for AMD MI300X GPUs.

Inference scaling benchmark of Qwen3.5-2B on AMD Instinct MI300X using ROCm and Hugging Face Transformers.

Agentic CUDA-to-ROCm migration loop - translates, compiles, profiles, and iteratively fixes GPU kernels with real hardware feedback.

An AI-powered command center that correlates AWS FinOps data with SecOps vulnerabilities using Fireworks AI and AMD Instinct MI300X.

Zero-config LLM benchmarking on AMD GPUs with ROCm. Auto-detect MI300X/MI250/Radeon, CUDA and CPU fallback.

Standalone AMD ROCm/PyTorch tools for pruning, expanding, and continued-pretraining an LLM checkpoint on a single MI300X GPU

Agentic AI store operations on AMD Instinct MI300X - Gemma tool-calling agents with enforced calculator grounding, human-in-the-loop approval, and a chat-first ops console. AMD ACT II Hackathon, Track 3.

White paper & reproducible benchmark suite for LLM inference optimization on AMD MI300X using ROCm 6.1

On-prem AI coding agent + LLM cost-router for regulated enterprises — 100% AMD, open-weight, air-gapped, Ed25519 tamper-evident audit. AMD ACT I 1st-place winner.

Custom HIP/ROCm kernel examples — matrix ops, attention, flash-attention optimized for AMD CDNA architecture

Multi-agent fine-tuning pipeline for Indonesian LLMs on AMD Instinct MI300X with ROCm

High-throughput LLM inference engine optimized for AMD Instinct MI300X GPUs

Production-ready LLM inference server optimized for AMD ROCm/MI300X. OpenAI-compatible API, PagedAttention, continuous batching, tensor parallelism, INT4/INT8 quantization.

GPU benchmarks for AMD MI300X — memory bandwidth, FLOPS, inference throughput across frameworks

Docker images for AMD ROCm development — PyTorch, TensorFlow, vLLM, Triton on MI300X

Meridian — An AI-first project management platform featuring a serverless architecture powered by React and an AMD MI300X Inference Endpoint.

Training-throughput benchmarks on AMD MI300X — LoRA, QLoRA, full SFT, DPO, GRPO with reproducible measurement

Self-contained fine-tuning recipes for AMD MI300X — LoRA, QLoRA, SFT, DPO, GRPO on ROCm 6.x

Adversarial LLM evaluation, powered by Gemma on AMD — AMD Developer Hackathon ACT II, Track 3