Microsoft has announced the 'Surface RTX Spark Dev Box' powered by NVIDIA RTX Spark, offering up to 1 PFLOPS of computing power for smooth local AI execution.



Microsoft has announced the ' Surface RTX Spark Dev Box, ' a developer PC equipped with

NVIDIA RTX Spark . Designed for AI development in a local environment, this PC boasts a maximum processing power of 1 PFLOPS (petaflops).

Building the next generation of devices for developers: Surface RTX Spark Dev Box | Microsoft Devices Blog
https://blogs.windows.com/devices/2026/06/02/building-the-next-generation-of-devices-for-developers-surface-rtx-spark-dev-box/

Surface RTX Spark Dev Box - YouTube














As AI models become more feature-rich and complex, and agent workflows require sustained computing power, cloud costs can be incurred with each iteration , even for tasks that don't require state-of-the-art AI models.

The Surface RTX Spark Dev Box is a developer PC developed to change this situation. With up to 1 petaflop of computing power, it brings powerful AI computing capabilities to the edge, allowing developers to limit the use of cutting-edge models to truly cutting-edge problems and handle the rest with their own hardware. This gives developers control over the allocation of their computing budget, enabling more efficient and responsive development workflows.

At the heart of the Surface RTX Spark Dev Box is the NVIDIA RTX Spark. The NVIDIA RTX Spark is a SoC announced by NVIDIA on June 1, 2026, and features a 1 petaflop Blackwell GPU, a 20-core Grace CPU, and 128GB of unified memory.

NVIDIA finally announces 'NVIDIA RTX Spark,' a SoC for laptops, featuring 128GB of memory, an NVIDIA GPU, and an Arm-based CPU for high-performance local AI execution - GIGAZINE



This provides sufficient computing power to run AI models with over 120 billion parameters locally and interactively within a 1 million token context, or to fine-tune AI models that previously required cloud GPU instances.

Furthermore, the Surface RTX Spark Dev Box features an aluminum chassis designed to function as a heatsink. This allows it to handle critical developer workloads such as long-running training jobs, large-scale model inference, and complex agent pipelines that require consistent and sustained performance.



The Surface RTX Spark Dev Box comes with Windows 11 Pro pre-configured at the developer level. This provides intentionally chosen default settings, pre-installed tools, and tuned settings, making the development environment available by default from the first sign-in.

This configuration is designed to help developers concentrate on their work and includes a dark theme, a simplified taskbar for development, removed widgets, Do Not Disturb mode enabled, and Developer Mode enabled. The default shell is PowerShell 7 .

WSL 2 is configured with GPU passthrough and CUDA support. Additionally, VS Code, GitHub Copilot, Git, Python, and Node.js are pre-installed. Whether in a Windows or WSL environment, your favorite IDEs, agents, coding assistants, frameworks, and libraries can all run on the Surface RTX Spark Dev Box.

Furthermore, security is a key consideration to make it a suitable choice for developers handling highly confidential AI models, proprietary data, and valuable intellectual property. Powerful GPUs and unified memory allow developers to store more AI models and intellectual property locally, enabling them to manage more AI models and data locally.

Regarding the appearance of the Surface RTX Spark Dev Box, some have pointed out that 'from a distance, it looks like a flattened Xbox Series X.'




Related Posts:

in AI,   Video,   Hardware, Posted by logu_ii