Microsoft Releases Windows Dev Kit “for Arm, on Arm”

October 27, 2022 by Aaron Carman

Hoping to create a one-stop shop for Arm-based AI development, Microsoft announced its Windows Dev Kit: hardware tools to develop for Arm, on Arm.

Microsoft has announced the release of the Windows Dev Kit 2023, formerly discussed under the code name Project Volterra in May 2022. Available to developers in eight countries (Australia, Canada, China, France, Germany, Japan, the U.K., and the U.S.), the Windows Dev Kit is built to allow developers to easily produce Windows apps for Arm, on Arm. Compared to traditional Arm development techniques, the Windows Dev Kit offers an easy-to-use, efficient environment for Arm development using familiar productivity tools in a compact and sleek form factor.


Windows Dev Kit

Windows Dev Kit in a typical development setting running tools for Windows app development. Image used courtesy of Microsoft


In this article, we'll review a brief history of Project Volterra, the key hardware at play for Windows Dev Kit 2023, and who, exactly, the new platform will benefit. 


From Project Volterra to Windows Dev Kit

Recognizing fragmentation in the world of computing hardware, Microsoft announced its intention to create a unified platform for developing Windows apps for Arm on an Arm-based architecture at Build in May 2022. This effort was named Project Volterra before eventually being released as the Windows Dev Kit.

From the project's onset, Microsoft wanted to break the longstanding tradeoff between flexibility and performance by integrating a high-powered, Arm-based computation unit with an end-to-end development toolchain purpose-built for Windows apps. Moreover, Microsoft has helped several open-source projects target Arm to further improve access to productivity tools while using the Windows Dev Kit.


A sample of the development tools that the Windows Dev Kit is designed to support

A sample of the development tools that the Windows Dev Kit supports. Image used courtesy of Microsoft


This has already put the Windows Dev Kit in a competitive position. When combined with its built-in AI capabilities and Microsoft’s Hybrid Loop architecture, the development platform may be a useful solution for designers looking to fast-track their Arm-targeted AI development workflow.


Purpose-built Hardware for AI

With artificial intelligence applications in mind, Microsoft carefully selected hardware to drive its Windows Dev Kit. At the center of the device is Qualcomm’s Snapdragon 8cx Gen 3 Compute Platform. When combined with the matching Neural Processing SDK, the platform gives developers a major head start on AI development right out of the box.


Qualcomm Snapdragon 8cx Gen 3 compute module

Qualcomm Snapdragon 8cx Gen 3 compute module in the Windows Dev Kit hardware. Image used courtesy of Microsoft


The computing platform also includes a neural processing unit (NPU), which gives the Dev Kit a sizeable advantage compared to traditional AI computing hardware. On some AI models, the NPU experienced an 80–90x speed increase compared to the CPU and approximately a 20x speed increase compared to the GPU. This not only allows the platform to rapidly complete complex AI tasks but frees up the CPU and GPU to accomplish their respective tasks, giving the unit more real-time versatility and responsiveness.

Finally, living up to its May 2022 promise, Microsoft has reported progress on launching a comprehensive toolchain for Arm app development to aid in the transition to the Dev Kit platform. Among the already available tools are the .NET Framework and Azure cloud services with more features expected to be available by the end of the year, including Visual Studio and VSCode.


Bringing Computation Together

The Windows Dev Kit marks a tremendous milestone in reunifying the computing hardware field. Instead of requiring developers to further complicate their workflow by using external debuggers or exporting binaries from an x64 machine, Arm development can be accomplished natively on an Arm machine, simplifying next-gen AI development using Windows.