Overview
PyTorch Conference North America is a premier event organized by the PyTorch Foundation and hosted by LF Events. Set to take place in San Jose, California, the conference gathers leading researchers, engineers, and practitioners from academia and industry to explore the future of open source AI and the PyTorch ecosystem. The event brings together researchers and developers to share cutting edge work, discuss practical applications, and collaborate on advancing AI technologies built around PyTorch and related tools.
Schedule and Agenda
Schedule at a glance
All times are listed in Pacific Daylight Time (PDT). The two-day program runs Tuesday, October 20, and Wednesday, October 21, with keynotes, technical sessions, and community components. The schedule includes keynote talks from prominent AI leaders, technical sessions spanning training, inference, and deployment, alongside hands-on workshops and a community expo. A poster track and poster sessions are also part of the program, culminating in social and networking opportunities.
Tuesday, October 20: The day features a mix of keynote sessions, technical talks, community expo activities, and poster sessions, capped by a social flare event to foster collaboration and conversation among attendees. Wednesday, October 21: The day continues with additional keynotes, deep dives on practical applications, and sessions across the full AI stack, including demonstrations of emerging tools and techniques.
Onsite Experience
The event emphasizes an on-site experience that supports collaboration and hands-on learning. Attendees can expect a traditional conference floor with a community expo, poster presentations, and technical tracks covering the breadth of PyTorch and allied technologies. There are opportunities for hands-on workshops, in-depth technical sessions, and discussions on responsible AI and kernel engineering, as well as sessions around operator performance, model optimization, and deployment across architectures.
Program Details and Tracks
The program highlights include keynote sessions by AI leaders, in-depth sessions on training and inference, and practical applications across the PyTorch ecosystem. Selected topics include vLLM, DeepSpeed, Ray, Helion, and Safetensors, with talks ranging from infrastructure and kernel engineering to agent-based systems and application-level demonstrations. The event also features a poster track and a poster session to showcase student and researcher work.
Who Should Attend
This conference is intended for researchers, developers, engineers, and data scientists who are actively building or leveraging PyTorch and related AI technologies. It is designed for people across the AI stack, from bare metal infrastructure to end-user applications and research contexts, and for those interested in staying abreast of the latest PyTorch developments and community innovations.
Resources and Logistics
The event is onsite and city-based in San Jose, CA, United States. Attendees may find venue and travel information on the official site, along with visa letter requests and accessibility considerations. The program includes CFP submissions and a schedule-at-a-glance page to help plan sessions of interest. A calendar add-to-service link is available for convenience. The conference also maintains a sponsor track with contributions from Google, Arm, Red Hat, Citadel Securities, Jane Street, Dell Technologies Capital, Clockwork, Fal, Lablup, Tigris Data, and other partners, reflecting a broad ecosystem support for PyTorch and open source AI.
FAQ Highlights (Visible on Page)
- What is the focus of the PyTorch Conference North America? The event centers on the future of open source AI and PyTorch ecosystem innovations.
- Where and when is the event held? San Jose, CA on October 20–21, 2026.
- What can attendees expect? Keynotes, technical talks, workshops, a community expo, posters, and networking opportunities.
In sum, PyTorch Conference North America is a deep-dive gathering for the open source AI community to learn, share, and shape the next generation of AI with PyTorch and related technologies. It provides a venue for researchers and practitioners to connect, collaborate, and explore practical and theoretical advances across the full AI stack.
