Canonical is once again proud to be one of the sponsors of the Nvidia GTC event! Happening virtually on November 8-11, the conference will feature a wide variety of sessions on AI, computer graphics, data science, and more.
During this GTC, Canonical will be hosting two sessions. Join for us a co-hosted speech with Nvidia on implementing a low-latency streaming solution with Android Cloud. And if you’re interested in data science, tune in to our talk on using open-source tools to build your MLOps infrastructure. Take a look at the details below and reserve your free spot for the talks today!
Canonical’s Sessions at GTC
Implementing a Low-latency Streaming Solution to Power Android Cloud Gaming [A31704]
Implementing an Android in the cloud solution involves various challenges including the development of a low latency streaming solution. In this talk, Canonical and NVIDIA collaborate towards building a platform to run Android in containers at high scale and density and implementing low latency video streaming powered by NVIDIA technology. This includes a zero-copy render-to-encode pipeline to keep latency on the server-side as minimal as possible. Enhanced with the power of video encoders on NVIDIA GPUs, high quality and low latency solutions are now possible. In this talk, we will walk jointly through the different aspects and challenges of building such a solution and provide insights into how NVIDIA technology makes this possible.
Speakers:
Chad Cooper, Product Manager for Cloud Gaming, NVIDIA
Simon Fels, Engineering Manager, Anbox, Canonical
Tailored MLOps infrastructure using Open Source tools [A31634]
MLOps setups are vastly different from one another. There is no single platform that can address all requirements. In order to have a cost-effective, secure, and fast MLOps platform you need to tailor build it. Let’s go over 5 real cases, and check how you can architect a solution from open source components. We will discuss MLOps setups for automatic customs declaration filling, 4000 4K video streams in parallel, multiple production lines in a car factory, deploying your models straight to NVidia Jetsons from MLOps pipeline, and more.
Speaker:
Maciej Mazur, Product Manager, Canonical