NVIDIA Unveils Nemotron 3.5 for Lightning-Fast On-Device AI in RTX Spark

NVIDIA has launched Nemotron 3.5 Bolt E4B, a compact language model for its RTX Spark platform, matching the response quality of much larger models while fitting within a 4GB memory budget. This innovation allows for seamless on-device AI interactions in gaming, eliminating cloud latency. RTX Spark now also supports native Windows on Arm development tools, expanding its ecosystem. Developers can access the model via NVIDIA's In-Game Inferencing SDK today.