Thin, Light and Built for Everything: NVIDIA RTX Spark Windows Laptops Gain New Models, Agent Tools and Games
NVIDIA expands RTX Spark laptops with new models, AI tools, and games, introducing speech models, agent resource management, and enhanced ray tracing support.
NVIDIA RTX Spark laptops, featuring an NVIDIA Blackwell RTX GPU and Grace CPU superchip with up to 128GB unified memory, are now available from ASUS, Dell, HP, Lenovo, Microsoft, and MSI. These systems deliver up to 1 petaflop of FP4 AI performance, integrating CUDA, RTX graphics, and personal AI supercomputing into premium Windows laptops designed for portability and versatility.
NVIDIA introduced Nemotron 3.5 Transcribe, an open-weight speech model that transcribes 40 languages, including seven Indic languages, and handles transcription, speaker labeling, translation, and audio queries via plain-language prompts. The model ranks No. 1 on Hugging Face Open ASR and No. 2 among open-weight models on Artificial Analysis, with approximately 4.3 billion active parameters and 80% fewer parameters than the top open-weight model, enabling efficient local deployment.
A new resource management feature in NVIDIA App allows users to prioritize GPU resources between foreground applications and background-running agents, optimizing performance based on immediate needs. The feature requires hardware-accelerated GPU scheduling in Windows Settings and is accessible via System > Advanced in the NVIDIA App.
Liquid AI’s LFM3 family introduces compact, fast models with a 1.7-billion-parameter base model outperforming its 2.6-billion-parameter predecessor, supporting local workflows like agent tasks and document processing. Developers can run LFM3 models on RTX Spark with native support through llama.cpp and vLLM, and fine-tune them locally for specialized applications.