Dynamo After Hours

​Overview

​Join us for a high-signal night with the NVIDIA developer community—built for developers pushing the limits of AI systems, inference, and scale.

​We’re bringing together teams behind some of the most exciting work across the stack for a fast-paced series of lightning talks (no fluff, all signal), followed by real conversations with builders like you and direct access to the developers behind the work. Whether you're optimizing throughput, experimenting with new frameworks, or scaling production workloads, this is where ideas and insights collide.

​Agenda:

• ​6:00 PM: Doors open

• ​6:30 PM: Lightning talks

• ​7:00 PM - 9:00 PM: Socialization with food and drinks

​ Lightning Talks Lineup (5 min each, 30 mins total)

• ​Microsoft Azure: Scaling inference together with NVIDIA Dynamo

• ​SGLang: Agentic inference

• ​LMCache: Improving transfer rates and fault tolerance with MP mode

• ​vLLM: WideEP fault tolerance

• ​Crusoe: Reducing TTFT by CPUMaxxing Tokenization

• ​EigenAI: From 11 to 25 #1-Speed Models: Scaling EigenInference with NVIDIA Dynamo

​Who Should Attend

​Developers working on LLMs, AI infra, and inference—whether you're tuning kernels, deploying at scale, or just getting your hands dirty.

​Why You Shouldn’t Miss This

​This is your chance to hear directly from the teams building the tools you’re using (or should be), get practical insights you can apply immediately, and connect with others solving the same hard problems.

​Spots are limited—register early.

​Resources and Legal

• ​Resources

​NVIDIA Privacy Policy: https://www.nvidia.com/en-us/about-nvidia/privacy-policy/