AMD posts async RL walkthrough on MI355X and benchmarks vs B300
AnushElangovan · x · 2026-08-19
AMD ROCm blog details running verl's fully asynchronous RL examples on AMD Instinct MI355X GPUs with ROCm.
- Hands-on: Covers Group Relative Policy Optimization (GRPO) on Qwen2.5-VL-7B and DAPO on Qwen2.5-Math-7B using Megatron and FSDP2 trainers.
- Benchmarks: Presents throughput numbers comparing AMD Instinct MI355X against the NVIDIA B300 for synchronous workloads.
More from Infra
- Anthropic's Multi-Level Monitoring for Astra Inference Revealed — AccBalanced · 2026-08-19
- Brex Report: Infrastructure Wins Over Apps in AI Hype — simonguozirui · 2026-08-19
- Alchemy Author: Not Just for Complex Projects — Simplest Way to Build Any Infra — samgoodwin89 · 2026-08-19
- Test: DeepSeek Harness achieves 99% cache hit rate with GLM and Kimi — sandyyevans · 2026-08-19
- 51WORLD Launches Embodied Data Infrastructure, Boosting Efficiency 10x — 量子位 · 2026-08-19
- Andrej Karpathy releases llm.c: Train LLMs in raw C/CUDA — goyalshaliniuk · 2026-08-19