Our vision for on-device AI and building best-in-class AI experiences at Microsoft Build 2023

Microsoft Build (aka MS Build or Build), a name synonymous with “developer event” since 2011, is the annual conference for developers using Microsoft’s evolving ecosystem of solutions. This year’s event, scheduled for May 23 to 25 in Seattle, WA., will offer virtual and in-person event experiences consisting of keynotes, technical deep dives, expert discussions and more. 

We are particularly excited to participate in Build to communicate the ability to move AI workloads to edge devices, including smartphones, cars, personal computers and Internet of Things (IoT) devices, to work together with the cloud to deliver more powerful, efficient and highly optimized artificial intelligence (AI). We’ll be showcasing several solutions, which you can learn about on our Build page, including:

• Qualcomm AI Stack: The Qualcomm AI Stack allows developers to build AI once and run it everywhere on devices powered by Snapdragon mobile platforms.
• Windows on Snapdragon: The next generation of mobile Windows computing on devices powered by our Snapdragon 8cx Gen3 Compute Platform.
• IoT — Qualcomm QCS6490: Purpose-built for IoT with support for multiple OSes and Wi-Fi 6E, the QCS6490 helps deliver powerful connections and reduced latency, along with accelerated AI and expanded interfaces for industrial and commercial use cases.

Let’s take a closer look at what we’ve got planned for the event.

 

 

Developer sessions

First up, we have four exciting sessions that you should attend:

• Qualcomm AI Stack for Developers: Vinesh Sukumar, Sr. Director of Product Management at Qualcomm Technologies, Inc., will discuss how our AI software enablement is helping developers develop AI models once, and then deploy them anywhere.
• Qualcomm Technologies and the power of Windows AI, Q&A: Jeff Gehlhaar, VP of Technology at Qualcomm Technologies, Inc., will discuss how developers can use the Snapdragon 8cx Gen3 Compute Platform, with its built-in neural processing unit (NPU), to build Windows AI experiences.
• Native apps for Windows on Snapdragon Compute Platforms: Rami Husseini, Director of Product Management at Qualcomm Technologies, Inc., will show the advantages of porting applications natively on Windows on Snapdragon compute platforms.
• On-device AI: Ankit Kapoor, Staff Product Management at Qualcomm Technologies, Inc., will showcase how developers can offload AI workloads and use AI accelerators to design their applications to use the full power and performance optimization of a Snapdragon compute platform.

 

 

Deeper dives and discussions

We’ll also have several deep dives and additional discussions you will want to check out:

• Leendert van Doorn, SVP of Engineering at Qualcomm Technologies, Inc., along with Microsoft executives, will provide a live Q&A session about our AI compute leadership.
• Megha Daga, Sr. Director of Product Management at Qualcomm Technologies, Inc., AI/ML Lead for IoT will present Edge AI in IoT — Defect Detection, in an on-demand, pre-recorded video.

 

 

Experience AI at our booth

You can also find us at our booth, where we’ll showcase several exciting demos and meetups:

• On-device stable diffusion on Windows on Snapdragon: Perform AI inferencing at the edge and in with a PC powered by Snapdragon 8cx Gen 3.
• AI-accelerated image processing: See how much faster a super resolution AI model runs on dedicated AI hardware versus the CPU.
• AI-accelerated game bots: Experience Air Derby with AI bots running on dedicated AI hardware on Snapdragon 8cx Gen 3.
• Native apps on Windows on Snapdragon: Learn how to build an app using native toolchains and see the improved performance of a native build versus emulated.
• Vision AI at the edge for industrial inspection: Showcases quality inspection on Windows 11 IoT Enterprise with AI acceleration on Qualcomm QCS6490.

 

 

Qualcomm Developer Portal

We are also pleased to provide an update that we are upgrading our developer experience with a new developer portal for Windows on Snapdragon. In it, you’ll find all sorts of learning resources to help you build or port apps to Windows on Snapdragon including documentation, tools and support. 

Start with our Getting Started guides which cover installation, project creation in Visual Studio and various methods to build. We’ve also included topics about the Qualcomm Neural Processing SDK for Windows so your apps can take advantage of the dedicated neural processing unit on our Windows on Snapdragon platform, and the Qualcomm Hexagon DSP. You can grab all the tools you need from our Tools page, and if you need additional help, post a question to our Discord channel. 

Be sure to revisit the Windows on Snapdragon Developer Portal frequently, as we continue to evolve its content.