PyTorch, Powering the Enterprise

Learn how to deploy and optimize AI workloads on Arm from cloud to edge through technical sessions, demos, and interactive experiences. Explore Gemma optimization, Google AI Edge and LiteRT, PyTorch and ExecuTorch, power-aware profiling, agentic AI, and physical AI. Hear from experts at Arm, Google DeepMind, Google, Renesas, and Nota AI. What to expect 💻 Explore cloud-to-edge AI workflows Explore AI technologies and workflows spanning cloud to edge, with technical sessions and demos from Arm engineers and ecosystem experts. ⚡ Try interactive AI demos and challenges Explore agentic and physical AI through interactive demos, and test your skills with developer challenges throughout the day. 🪪 Hack your Arm-powered digital badge Get your own Arm-powered digital badge to personalize and hack during the event. 🤝 Connect with engineers and developers Ask questions and connect directly with Arm engineers, ecosystem partners, and other developers. Agenda highlights Running Gemma open models on-device Ian Ballantyne, Google DeepMind Google AI Edge: Proven Local AI from Gemma to IoT Matthias Grundmann, Google From PyTorch to product: Accelerating edge AI through open ecosystems Kaushal Vora, Renesas Visual debugging and optimization with Google Model Explorer and MLIA Charlie Stokes, Arm Hands-on AI model optimization for Arm Tae-Ho Kim, NetsPresso® by Nota AI Power-aware PyTorch profiling: Bringing Arm telemetry into PyTorch for efficient cloud-edge inference Sarath Nandu R, Arm Agentic AI on Arm Neoverse CPUs in the cloud Na Li, Arm View the full agenda Arm Create After Hours | 5:30–9:00 PM PDT Stay for an Oktoberfest-inspired evening at The Tech Interactive. Personalize your Arm-powered digital badge, create a custom-printed T-shirt, and grab a drink. Connect with fellow developers over music and interactive experiences. Arm-powered digital badges are available exclusively to Arm Create daytime session attendees (10:00 AM–5:00 PM PDT), while supplies last.

Join dstack, Crusoe, and SGLang for an AI Infra Night in San Jose during PyTorch Conference. After a full day at the conference, come unwind with GPU experts, ML engineers, AI researchers, and AI infra enthusiasts working on GPUs, training, inference, and open-source AI systems. Expect drinks, food, and plenty of time to talk shop, compare notes on what you saw at the conference, and meet the dstack and Crusoe teams. Schedule 6:30 PM: Doors open, drinks, and snacks 7:00 PM – 9:30 PM: Networking, food, and drinks Additional information A short walk from the San Jose Convention Center. Exact address shared on registration. Space is limited. Registration required, first come, first served. Your hosts: dstack is an open-source orchestration stack for heterogeneous AI compute. It standardizes how development, training, and inference run across GPU clouds, Kubernetes, and on-prem clusters, on NVIDIA, AMD, and other silicon. Crusoe is a vertically integrated AI cloud built for the demands of production AI. From managed inference endpoints to GPU clusters, Crusoe gives teams the compute, managed services, and enterprise support they need to train and deploy AI. Disclaimer: Must be 21+ to attend, valid ID required. Capacity is capped. First come, first served. By signing up and/or attending, you consent to being photographed and may also receive emails.