NVIDIA Details GPU Sharing Controls and Gaming Updates for RTX Spark Laptops
NVIDIA is expanding the software offering around its RTX Spark Windows laptops, with GPU resource controls, local AI tools and gaming updates outlined in an October 28 company announcement. NVIDIA says systems are available from ASUS, Dell, HP, Lenovo, Microsoft and MSI.
The laptops combine a Blackwell RTX GPU and Grace CPU in a single chip, with configurations offering up to 128GB of unified memory. NVIDIA lists a starting price of $2,599, although its announcement covers the platform rather than detailing individual manufacturers’ configurations.
Sharing GPU resources with background agents
A resource management feature in NVIDIA App lets users choose how GPU resources are divided between foreground applications and agents running in the background. According to NVIDIA, the controls appear under System > Advanced and require Windows hardware-accelerated GPU scheduling to be enabled.
The company also introduced Nemotron 3.5 Transcribe, an open-weight speech model for transcription, speaker attribution, translation and questions about recorded audio. NVIDIA says it transcribes 40 languages. The model is available on Hugging Face, while quantized versions intended for local deployment are still described as coming soon.
RTX Remix extends its reach
On the gaming side, NVIDIA says RTX Remix 1.6 adds native RTX Spark support and allows Minecraft Java Edition 1.21.1 to run through the modding platform with path tracing and DLSS. The update also includes material-generation workflows for modders.
NVIDIA says Portal with RTX and the Half-Life 2 RTX demo receive RTX Spark support alongside rendering upgrades. Its announcement also lists ray tracing and DLSS 4.5 for S-GAME’s Phantom Blade Zero.
The practical focus is flexibility: giving users control over competing GPU workloads while broadening the software supported by the laptop platform. The announcement’s hardware and software capability claims come from NVIDIA, rather than independent testing.