NVIDIA and Microsoft Bring RTX Spark Laptops and Agent Security to Windows

Windows gets an agent security layer

At the Windows AI and Surface event at Dogpatch Studios, Nadella and Huang spoke in a fireside chat hosted by Sriram Krishnan. Huang tied NVIDIA's history to Windows. "If not for Windows there would be no GeForce," he said, according to NVIDIA's account of the event.

Pavan Davuluri, Microsoft's EVP of Windows and Devices, said Microsoft announced general availability of Microsoft Execution Containers (MXC). Per the source, MXC is OS-level infrastructure that lets agents run safely and persistently in the background, under operating system control. "We are building these primitives directly into Windows, so agents can be secured, observed and governed," Davuluri said. He named MXC, Microsoft Security and Agent 365 as the pieces behind that approach.

Nadella said the goal was to make "the desktop the most secure place for agents to execute." Huang compared the step to earlier platform shifts: "Just as Windows and DirectX revolutionized how applications were built, MXC is going to revolutionize how agents are built and deployed."

RTX Spark: local AI in laptops and compact desktops

RTX Spark puts the NVIDIA AI stack into Windows laptops and compact desktops. According to NVIDIA, laptop preorders are open and the machines are available October 16, while compact desktops go on sale in November. Systems are coming from Acer, ASUS, Dell, HP, Lenovo, Microsoft, MSI and Gigabyte. Microsoft also introduced Surface Laptop Ultra, built around RTX Spark.

NVIDIA lists the following specifications:

• An NVIDIA Blackwell RTX GPU with up to 6,144 cores.
• An up to 20-core NVIDIA Grace CPU connected at 600 GB/s.
• One petaflop of FP4 AI performance and up to 128GB of unified memory.

Davuluri said that with this memory and compute, "you can run models on this laptop that simply don't fit on a traditional machine." NVIDIA says RTX Spark can run Qwen 3.8 Flash Next, a 125B model with 51B n-gram that, in the company's words, matches the intelligence of many cloud models, unmetered and without sending data to the cloud.

The platform runs the full NVIDIA CUDA platform, so developers can move models, tools and workflows without rewriting them, using the same stack from RTX Spark to DGX Station. For creators, NVIDIA cites 5th-generation Tensor Cores with NVFP4 support, hardware-accelerated AV1, and 4:2:2 video encode and decode. For gamers, it says AAA games can run at 1440p over 100 frames per second with DLSS 5, Reflex and G-SYNC. The compact desktop version uses the same superchip in a small chassis designed for 24/7 operation, aimed at keeping agents running continuously.

DGX Station comes to Windows

The event also previewed NVIDIA DGX Station for Windows, which NVIDIA describes as the first deskside AI supercomputer to bring GB300 Grace Blackwell-class AI infrastructure into the Windows ecosystem. Until now, DGX Station ran on Linux, which forced enterprise developers to maintain separate environments: Linux for heavy AI workloads and Windows for productivity tools and applications. NVIDIA notes that the vast majority of Fortune 500 companies are standardized on Windows.

The system runs on the GB300 Grace Blackwell Ultra Desktop Superchip and offers 748GB of coherent memory and up to 20 petaFLOPS of FP4 AI compute, which NVIDIA says is enough to run models up to a trillion-parameter scale locally. Teams can build always-on agents that connect to existing Windows applications and infrastructure, and fine-tune and run large models without leaving their main machine. Linux AI toolchains remain available through WSL. NVIDIA invites interested users to sign up for availability news, and no release date was given.

What it means for users and developers

For buyers, the announcement sets concrete dates: RTX Spark laptops from October 16 and compact desktops in November. For enterprises and developers, the pitch is to run capable models and agents on a Windows machine, with Microsoft's MXC providing the control and security layer. Huang described the shift in simple terms: when the agent is on your computer, "it is now your personal assistant."