Microsoft has announced the Surface RTX Spark Dev Box, a compact workstation for AI developers. Powered by an NVIDIA RTX Spark superchip with 128GB of unified memory, it delivers up to 1 petaflop of local compute for training large models. The device ships with a pre-configured Windows 11 Pro environment, aiming to streamline workflows and reduce reliance on cloud services for prototyping.
Surface RTX Spark Dev Box
- ▪The device is intended for local-first AI development, allowing developers to prototype, fine-tune, and run models on their desk
- ▪The Surface RTX Spark Dev Box will be available later in 2026 in the U.S. exclusively on Microsoft.com
- ▪The Surface RTX Spark Dev Box has an aluminum chassis that also functions as a heatsink
- ▪The device is a direct replacement for Qualcomm’s canceled Snapdragon Dev Kit, which was originally planned to ship two years ago
- ▪Microsoft announced the Surface RTX Spark Dev Box, a compact PC designed for developers
Developer workflow transformation drivers
- ▪The Dev Box allows developers to reserve cloud-based "frontier model" calls for the most complex problems by handling other tasks locally
- ▪Modern software development involves longer-running jobs, larger models, and a need for local prototyping to manage cloud costs
NVIDIA RTX Spark superchip specifications
- ▪The device features 128GB of unified memory and has a 100-watt thermal envelope
- ▪The Surface RTX Spark Dev Box is powered by an Nvidia Arm-based RTX Spark superchip, which combines a Blackwell RTX GPU and a Grace CPU
- ▪The superchip can deliver up to 1 petaflop of AI compute, enough to run models with over 120 billion parameters locally
Pre-configured Windows development environment
- ▪Preinstalled tools on the device include Visual Studio Code, GitHub Copilot, Git, Python, and Node.js
- ▪The Dev Box ships with Windows 11 Pro pre-configured for developers, with settings like dark theme enabled and widgets removed
- ▪Developer Mode is enabled by default, PowerShell 7 is the default shell, and WSL 2 is configured with GPU passthrough and CUDA support
Secured-core PC architecture
- ▪The device is built on a Secured-core PC architecture aligned with Microsoft’s Zero Trust security principles
- ▪Security features include BitLocker encryption and Microsoft Defender protection, with support for Entra ID and Intune for organizational management
Microsoft AI stack integration
- ▪The AI Toolkit for VS Code is included to bring model conversion, fine-tuning, and evaluation into the editor
- ▪The device integrates with Windows ML with TensorRT, Windows Copilot Runtime, and Microsoft Foundry to connect local prototyping with production deployment
Story comments
Loading comments…