NVIDIA
NVIDIA’s Edge Model Hits 27.7 Tokens/s, but Doubled Speed Still Faces Limits
A developer hit 27.7 tokens/s with Nemotron 3.5 Bolt E4B on Jetson Orin Nano Super, 1.8× NVIDIA’s initial result.
1 article tagged with this topic
A developer hit 27.7 tokens/s with Nemotron 3.5 Bolt E4B on Jetson Orin Nano Super, 1.8× NVIDIA’s initial result.