Skip to content
Sıfırıncı Dakika
Latest

NVIDIA RTX Spark laptops gain new models, agent tools and games

HardwareNew1 min read
NVIDIA Blog

In brief

NVIDIA announced the open-weight Nemotron 3.5 Transcribe speech model and Liquid AI's LFM3 model family for RTX Spark Windows laptops. A new NVIDIA App feature also lets users control how GPU resources are shared between apps and background agents. Phantom Blade Zero and a Minecraft RTX Remix 1.6 update are also on the way.

  • Nemotron 3.5 Transcribe: open-weight speech model supporting 40 languages
  • GPU resource sharing between agents and apps in NVIDIA App
  • Day-zero support for Liquid AI's LFM3 model family
  • Ray tracing and DLSS 4.5 in Phantom Blade Zero
  • Path tracing and DLSS in Minecraft via RTX Remix 1.6

RTX Spark laptops now available

NVIDIA said RTX Spark laptops, which combine a Blackwell RTX GPU and a Grace CPU in a single superchip, are now sold by ASUS, Dell, HP, Lenovo, Microsoft and MSI. The systems offer up to 128GB of unified memory and up to 1 petaflop of FP4 AI performance.

Nemotron 3.5 Transcribe

NVIDIA's new open-weight speech model transcribes recorded conversations, labels speakers and translates. It supports 40 languages and can handle all three tasks with a single prompt. It is available on Hugging Face in BF16, FP8 and NVFP4 formats.

GPU resource management for agents

A new feature in NVIDIA App lets users choose how GPU resources are shared between foreground apps and background AI agents. It requires hardware-accelerated GPU scheduling to be enabled in Windows Settings.

Liquid AI LFM3 support

Liquid AI's LFM3 family, which spans text, vision, audio and decision models on one shared backbone, gets day-zero support on RTX Spark. According to the company, its new 1.7-billion-parameter base model outperforms its previous 2.6-billion-parameter model.

Games and other updates

S-GAME's Phantom Blade Zero launches with ray tracing and DLSS 4.5. RTX Remix 1.6, which brings path tracing and DLSS to Minecraft, is now available for RTX Spark.

Why it matters

As local AI and agents become more common, NVIDIA's open-weight speech model and GPU resource controls are a notable step for those running AI on laptops.

Sources

Related stories