Nemotron 3.5 Bolt E4B Hits 27.7 tok/s on Jetson Orin Nano After Optimization

The Embedl team optimized NVIDIA's Nemotron 3.5 Bolt E4B to run at 27.7 tokens per second on the $399 Jetson Orin Nano Super (8GB), 1.8x faster than llama.cpp, making local LLM inference more practical on small devices.

2026-11-05 ~ 2026-11-06 · 2 related posts

• Nemotron 3.5 Bolt E4B hits 27.7 tok/s on Jetson Nano, 1.8x faster than llama.cpp — Course_Latter · 2026-11-05
• Nemotron 3.5 Bolt E4B hits 27.7 tok/s on Jetson Orin Nano Super, 1.8x faster than baseline — Any_Frame9721 · 2026-11-06