Panel discussion - 40 minutes
This Birds of a Feather (BoF) session invites developers, architects, and technologists to collaborate on advancing AI and ML development on Snapdragon-powered Windows devices. With on-device AI gaining traction for efficiency and privacy, this session will explore key opportunities and challenges developers face in this evolving space. Topics of discussion include: Best practices for optimizing AI applications for Windows on Snapdragon. Developer tools and frameworks: What works, what’s missing, and how they can improve. Insights into leveraging on-device AI for privacy, power efficiency, and performance. Community collaboration: How Qualcomm, LMStudio, and Ollama can better support the developer ecosystem. Join us to share your experiences, brainstorm solutions, and shape the future of AI and ML development on Snapdragon platforms. Whether you’re an experienced developer or just getting started, your insights can help pave the way for new innovations.
Technical presentation - 30 minutes (including q&a)
As ARM64 adoption grows across PCs and embedded devices, developers need robust tooling to build, optimize, and debug applications efficiently. This session explores the latest advancements in developer tools, compilers, debuggers, emulation strategies, and profiling techniques that simplify development on ARM64. We will discuss how modern toolchains support cross-compilation, how to optimize performance for ARM64-native applications, and best practices for ensuring seamless development workflows across multiple architectures. Attendees will gain insights into how development environments, open-source frameworks, and performance analysis tools are evolving to make ARM64 development as seamless as possible.
Tutorial - 60 minutes
The evolution of on-device AI is transforming how developers build and deploy machine learning models, enabling real-time performance, enhanced privacy, and power-efficient execution. Windows on Snapdragon (WoS) provides a robust AI development ecosystem, allowing developers to run AI models directly on the device without relying on cloud inference. This session will explore best practices for optimizing AI workloads, leveraging hardware acceleration, and utilizing industry-standard frameworks to deploy AI models efficiently on WoS-powered devices. Attendees will gain insights into the AI developer workflow, from model conversion and quantization to execution and optimization, ensuring seamless performance for a wide range of AI applications.
Dileep Karpur is a Staff Software Product Manager at Qualcomm Technologies, Inc, where he works as part of the Windows on Snapdragon team. He's responsible for leading the development and management of tools and APIs tailored for developers, executing developer-centric initiatives, and fostering an engaged developer community.