RTX 5090 Demo: Running Qwen 3.8 27B Locally at 115 tok/s
rohanpaul_ai · x · 2026-08-17
Rohan Paul shared a demo of running Qwen 3.8 27B locally on a system with an RTX 5090 (32GB VRAM), achieving a speed of 115 tokens/sec. Notably, the model's official BF16 checkpoint is 55.6 GB, suggesting the demo likely utilized a quantized version to fit within the 32GB memory limit.
Related event: Qwen3.8-27B Benchmarked Across GPUs: Local Deployment Proves Viable(7 posts)→
More from Embodied
- Robotics RL tutorial: deploying to real hardware and sim-to-real challenges — ShawnHymel · 2026-08-17
- Gravis Robotics raises $200M led by SoftBank to automate excavators — lukas_m_ziegler · 2026-08-17
- Stone carving boss: Robots didn't kill jobs, they revived the craft — Afinetheorem · 2026-08-17
- Polymarket gives Tesla's Optimus release by year-end an 11% chance — Polymarket · 2026-08-17
- US robotics company Foundation pitches humanoid robots for southern border patrol — Polymarket · 2026-08-17
- US-China Robotics Decoupling Begins: FCC Ban Reshapes Supply Chain — Rewkang · 2026-08-17