At a glance
- NVIDIA RTX Spark powers thin and light laptops built for everything that developers, creators and gamers do
- RTX Spark laptops are now available from ASUS, Dell, HP, Lenovo, Microsoft and MSI
- These powerful systems combine an NVIDIA Blackwell RTX GPU and an NVIDIA Grace CPU in a single superchip with up to 128GB of unified…
NVIDIA Expands RTX Spark Laptop Lineup with Agent Tools and AI Models
NVIDIA’s RTX Spark platform is extending its reach across the premium Windows laptop market, with new models now available from ASUS, Dell, HP, Lenovo, Microsoft, and MSI. The systems pair an NVIDIA Blackwell RTX GPU with an NVIDIA Grace CPU in a single superchip, delivering up to 128GB of unified memory and up to 1 petaflop of FP4 AI performance. NVIDIA positions the design as a personal AI supercomputer in a thin-and-light form factor, bringing the CUDA platform and RTX graphics capabilities to mobile workstations.
Nemotron 3.5 Transcribe Targets Local Audio Processing
A central software addition is NVIDIA Nemotron 3.5 Transcribe, an open-weight speech model introduced for transcription, speaker attribution, and translation. The model handles 40 languages, including seven Indic languages, and can be guided through plain-language prompts such as requesting speaker labels, translating audio into English, or extracting decisions from a recorded meeting.
NVIDIA reports that Nemotron 3.5 Transcribe ranks No. 1 on the Hugging Face Open ASR leaderboard and No. 2 among open-weight models on Artificial Analysis. The model uses approximately 4.3 billion active parameters, which NVIDIA states is about 80% fewer than the No. 1 open-weight model. Quantized versions are planned to enable local execution on RTX Spark and DGX Spark systems, Jetson AGX Thor devices, and NVIDIA Blackwell data center GPUs. NVFP4 versions will be deployable through NeMo-Speech.cpp or vLLM.
Liquid AI’s LFM3 Family Gains Day-Zero Support
Liquid AI’s LFM3 foundation model family is also arriving with day-zero support on RTX Spark. The family spans text, vision, audio, omni, and decision models built on a shared backbone, with base and post-trained models available as open weights. Liquid AI’s 1.7-billion-parameter base model is described as outperforming the company’s previous 2.6-billion-parameter base model.
The models are intended for local agent workflows, including inbox triage, calendar management, document and screen reading, speech transcription and translation, spoken questions about handwritten math, and support ticket routing. Developers can run LFM3 models on RTX Spark using llama.cpp and vLLM, with local fine-tuning supported.
GPU Resource Controls for Agents and Foreground Apps
NVIDIA App is receiving a resource management feature that lets users allocate GPU compute between foreground applications and background agents. Users can prioritize the active application for maximum performance or assign more resources to a background-running agent to accelerate task completion. The setting is accessible through System > Advanced in the NVIDIA App and requires hardware-accelerated GPU scheduling to be enabled in Windows Settings.
Gaming and Creative Updates: DLSS 4.5 and RTX Remix 1.6
S-GAME’s action role-playing game Phantom Blade Zero launches on October 28 at 7 p.m. PT on Steam and the Epic Games Store. The title includes ray-traced effects developed with NVIDIA, adding bounced lighting, detailed reflections, and realistic shadows on foliage and terrain. On RTX Spark laptops and GeForce RTX-powered PCs, DLSS Dynamic Multi Frame Generation provides up to 6x frame generation, with an adjustable multiplier to target specific frame rates.
RTX Remix, NVIDIA’s modding platform for remastering classic games, has reached version 1.6. The update brings path tracing and DLSS to Minecraft Java Edition 1.21.1, alongside DLSS 5 with 3D-Guided Neural Rendering, up to 6x Multi Frame Generation, native RTX Spark support, and agent-assisted workflows. Portal with RTX and the Half-Life 2 RTX demo also receive RTX Remix support, including a new Performance Mode for DLSS Frame Generation and options to accelerate path tracing.
The RTX Remix 1.6 release includes a redesigned ComfyUI workflow for generating physically based materials directly in the application, batch processing for full projects, and presets aimed at less technical creators. NVIDIA agent skills allow compatible AI agents to work directly with RTX Remix to speed up remastering projects.
Design and Mobility Specifications
RTX Spark laptops are available in multiple screen sizes, colors, touchscreen configurations, pen-support options, and two-in-one convertible designs. NVIDIA states the systems can be as thin as 14mm and as light as three pounds, with all-day battery life. The slim profiles are enabled by placing the GPU, CPU, and up to 128GB of unified memory on a single power-efficient superchip.
For data center and infrastructure operators, the RTX Spark announcement signals continued pressure to support AI workloads at the edge, with unified memory architectures and local model execution reducing reliance on cloud inference for transcription, agent tasks, and creative rendering.
MegaWatt Intelligence Desk
Hyperscale & Grid Analyst. `More stories by $MegaWatt Intelligence Desk →`
