At a Microsoft Windows AI and Surface event in San Francisco on October 7, NVIDIA CEO Jensen Huang and Microsoft CEO Satya Nadella announced hardware and software work to support AI agents on Windows PCs. Microsoft said it is building security controls into Windows, while NVIDIA introduced RTX Spark systems for running AI models locally.

Microsoft announced general availability of Microsoft Execution Containers (MXC), operating-system infrastructure that lets agents run persistently in the background under Windows control. Microsoft said the tools are designed to secure, observe and govern agents, alongside Microsoft Security and Agent 365. Nadella said the desktop needed to become a secure place for agents to execute; Huang described MXC as a foundation for building and deploying agents.

RTX Spark puts NVIDIA’s AI software stack into Windows laptops and compact desktops. Laptop preorders opened October 7, with availability set for October 16; compact desktops are planned for November. Microsoft’s Surface Laptop Ultra will use RTX Spark and offer up to 128GB of unified memory and up to one petaflop of AI compute. Systems are also planned from Acer, ASUS, Dell, HP, Lenovo, Microsoft, MSI and Gigabyte, in laptop and compact desktop designs intended for agents that run continuously.

RTX Spark combines an NVIDIA Blackwell RTX GPU with up to 6,144 cores and an NVIDIA Grace CPU with up to 20 cores, connected at 600 GB/s. It delivers one petaflop of FP4 AI performance and up to 128GB of unified memory. NVIDIA said it can run Qwen 3.8 Flash Next, a 125B model with 51B n-gram, locally without metered use or sending data to the cloud. Developers can carry over models, tools and workflows from RTX Spark to DGX Station using the same CUDA platform. Creator features include 5th-generation Tensor Cores with NVFP4 support, AV1 acceleration, 4:2:2 video encoding and decoding, DLSS and RTX ray tracing. NVIDIA said gamers can play AAA games at 1440p and over 100 frames per second with DLSS 5, Reflex and G-SYNC.

NVIDIA also previewed DGX Station for Windows, a deskside AI supercomputer based on the GB300 Grace Blackwell Ultra Desktop Superchip. It provides 748GB of coherent memory and up to 20 petaFLOPS of FP4 AI compute, which NVIDIA says can run models at up to a trillion-parameter scale locally. The system is intended for developers and researchers to build always-on agents connected to Windows apps and infrastructure, and to fine-tune and run inference on large models without leaving their main machine. Linux AI toolchains remain available through WSL. Before this announcement, DGX Station ran on Linux; NVIDIA said Windows-standardized Fortune 500 companies had faced separate environments for AI workloads and everyday productivity, with developers either switching to Linux or working with limited hardware in Windows.

At a Microsoft Windows AI and Surface event in San Francisco on October 7, NVIDIA CEO Jensen Huang and Microsoft CEO Satya Nadella announced hardware and software work to support AI agents on Windows PCs. Microsoft said it is building security controls into Windows, while NVIDIA introduced RTX Spark systems for running AI models locally.

Microsoft announced general availability of Microsoft Execution Containers (MXC), operating-system infrastructure that lets agents run persistently in the background under Windows control. Microsoft said the tools are designed to secure, observe and govern agents, alongside Microsoft Security and Agent 365. Nadella said the desktop needed to become a secure place for agents to execute; Huang described MXC as a foundation for building and deploying agents.

RTX Spark puts NVIDIA’s AI software stack into Windows laptops and compact desktops. Laptop preorders opened October 7, with availability set for October 16; compact desktops are planned for November. Microsoft’s Surface Laptop Ultra will use RTX Spark and offer up to 128GB of unified memory and up to one petaflop of AI compute. Systems are also planned from Acer, ASUS, Dell, HP, Lenovo, Microsoft, MSI and Gigabyte, in laptop and compact desktop designs intended for agents that run continuously.

RTX Spark combines an NVIDIA Blackwell RTX GPU with up to 6,144 cores and an NVIDIA Grace CPU with up to 20 cores, connected at 600 GB/s. It delivers one petaflop of FP4 AI performance and up to 128GB of unified memory. NVIDIA said it can run Qwen 3.8 Flash Next, a 125B model with 51B n-gram, locally without metered use or sending data to the cloud. Developers can carry over models, tools and workflows from RTX Spark to DGX Station using the same CUDA platform. Creator features include 5th-generation Tensor Cores with NVFP4 support, AV1 acceleration, 4:2:2 video encoding and decoding, DLSS and RTX ray tracing. NVIDIA said gamers can play AAA games at 1440p and over 100 frames per second with DLSS 5, Reflex and G-SYNC.

NVIDIA also previewed DGX Station for Windows, a deskside AI supercomputer based on the GB300 Grace Blackwell Ultra Desktop Superchip. It provides 748GB of coherent memory and up to 20 petaFLOPS of FP4 AI compute, which NVIDIA says can run models at up to a trillion-parameter scale locally. The system is intended for developers and researchers to build always-on agents connected to Windows apps and infrastructure, and to fine-tune and run inference on large models without leaving their main machine. Linux AI toolchains remain available through WSL. Before this announcement, DGX Station ran on Linux; NVIDIA said Windows-standardized Fortune 500 companies had faced separate environments for AI workloads and everyday productivity, with developers either switching to Linux or working with limited hardware in Windows.