AMD Releases Instella-MoE-16B: A Fully Open Mixture-of-Experts LLM
mpuchala · reddit · 2026-08-03
AMD has announced the release of Instella-MoE-16B-A3B, a Mixture-of-Experts (MoE) large language model with 16 billion total parameters and 2.8 billion active parameters. Trained on AMD Instinct GPUs, the model is fully open-source, aiming to empower the AI community and offer a cost-effective inference alternative.
Related event: AMD Launches Fully Open-Source Instella-MoE-16B Model(2 posts)→
More from Infra
- NVIDIA RTX 50 Series GPU Prices Surge Up to 30% in Korea Amid Memory Cost Hikes — basedjensen · 2026-08-03
- MiniMax H3 Local Test: 10s 1080p Video Takes 58 Minutes on RTX 5090 — koakoAI · 2026-08-03
- Running MiniMax-H3 on Nvidia B200: Generates 15s 2K Video — tsi_org · 2026-08-03
- Intel Accelerates Fab Construction, AMD and Storage Earnings to Test AI Demand — 创业邦 · 2026-08-03
- Unsloth Founder Validates: Qwen3-8-27B Runs on Just 17GB VRAM — quantier · 2026-08-03
- Cutting LLM Context Costs: Developers Pack Text into PNGs for Fable 5 — rohanpaul_ai · 2026-08-03