Developers Blogs#
Primus Projection: Estimate Memory and Performance Before You Train
Learn how to use the Primus projection tool to estimate memory and performance for large-scale LLM training on AMD Instinct™ accelerator platforms.
Getting Started with FlyDSL Nightly Wheels on ROCm
A practical guide to installing and using FlyDSL nightly wheels on ROCm for fast, Python-native GPU kernel development
Customizing Kernels with hipBLASLt TensileLite GEMM Tuning - Advanced User Guide
Master hipBLASLt TensileLite Tuning. Learn to build custom kernels that deliver 150%-250% faster GEMM performance on AMD Instinct™ MI300X GPUs
Training a Robotic Arm Using MuJoCo and JAX on AMD Hardware with ROCm™
A complete guide to training an RL-based pick-and-lift robotic arm in MuJoCo with JAX, running on AMD hardware via ROCm.
Edge-to-Cloud Robotics with AMD ROCm: From Data Collection to Real-Time Inference
This blog demonstrates a comprehensive Edge-to-Cloud robotics AI solution powered by the AMD ecosystem and the Hugging Face LeRobot framework.
hipBLASLt Online GEMM Tuning
Learn how to improve model performance with hipBLASLt online tuning merged into LLM framework
GROMACS Performance on AMD Instinct MI355X
Explore GROMACS molecular dynamics performance benchmarks on AMD Instinct MI355X GPUs with HIP acceleration.
Agentic Diagnosis for LLM Training at Scale
Explore how AI agents diagnose LLM training incidents — from RCCL hangs to throughput regressions — in one prompt with MaxText-Slurm.
Getting Started with ComfyUI on AMD Radeon™ RX 9000 Series GPUs
Learn how to set up and optimize ComfyUI on AMD Radeon RX 9000 GPUs with ROCm 7.1 — solve common issues and start generating.
MaxText-Slurm: Production-Grade LLM Training with Built-In Observability
MaxText-Slurm: A unified launch system for production-grade LLM training with observability on AMD GPU clusters.
LuminaSFT: Generating Synthetic Fine-Tuning Data for Small Language Models
Learn how task-specific synthetic data can improve small language model performance and explore results from the LuminaSFT study.
Primus-Pipeline: A More Flexible and Scalable Pipeline Parallelism Implementation
Learn how to use our flexible and scalable pipeline parallelism framework with Primus backend and AMD hardware.