NVIDIA unveils RTX Spark for the AI-powered PC era
A new superchip aims to transform personal computing by combining powerful AI performance, enhanced security and Windows-native agent experiences.
NVIDIA and Microsoft have introduced RTX Spark, a new Windows PC platform designed for personal AI agents.
NVIDIA describes RTX Spark as a 1-petaflop superchip that combines its AI and graphics stack with Windows-native agent capabilities.
NVIDIA’s Blackwell architecture powers the platform and supports up to 128GB of unified memory.
According to NVIDIA, RTX Spark will allow users to run local AI agents, large language models, creative workflows and advanced games on laptops and compact desktop PCs.
The company said the platform can run 120-billion-parameter large language models with up to 1 million tokens of context locally.
NVIDIA and Microsoft are also introducing new Windows security primitives and NVIDIA OpenShell to help agents run securely on primary devices.
OpenShell will allow users to define what agents can and cannot do, route queries to local models according to privacy policies and mask personal information when cloud models are used.
RTX Spark-powered laptops and compact desktops are expected to be available this autumn from manufacturers including ASUS, Dell, HP, Lenovo, Microsoft Surface and MSI, with Acer and GIGABYTE models to follow.
Why does it matter?
RTX Spark reflects the industry shift towards AI-native personal computers, where more AI processing happens locally on the device. Running agents and large models on PCs could improve privacy, reduce latency and make advanced AI tools less dependent on cloud access. The governance question is whether local agents can operate with clear user permissions, strong containment and meaningful accountability as they gain the ability to search files, interact with apps and execute tasks across a personal device.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our chatbot!