Leveraging Nvidia's Blackwell for Efficient Inference of Large Language Models

As large language models (LLMs) continue to grow in size and complexity, the demand for efficient inference capabilities becomes paramount. Models like LLama 3.3 405B and DeepSeek-R1, with their billions of parameters, pose significant challenges in terms of computational resources and energy consumption. In this talk, we will explore how Nvidia's latest GPU architecture, Blackwell, is designed to address these challenges.

Inference at Scale: Kubernetes and NVIDIA for AI Workloads

As AI workloads continue to grow, the challenge of deploying inference at scale has become critical. In this session, we’ll explore how Kubernetes and NVIDIA GPUs work in tandem to deliver scalable, efficient, and reliable AI inference services. Discover deployment patterns, best practices, and real-world strategies to optimize GPU utilization and performance, all within the Kubernetes ecosystem. Whether you’re running deep learning models or complex analytics, you’ll gain practical insights to supercharge your inference workloads and meet the demands of modern AI applications.

Accelerating AI Inference with Kubernetes and NVIDIA GPUs

Join us as we explore how to harness the power of Kubernetes and NVIDIA GPUs to accelerate AI inference workloads. In this session, we'll cover best practices for deploying inference as a service, leveraging GPU resources efficiently, and scaling AI applications seamlessly. Learn about cutting-edge tools, deployment patterns, and real-world use cases that demonstrate how Kubernetes and NVIDIA GPUs work together to deliver high-performance inference solutions.

Learn and Apply Windows ML and Azure ML in your Awesome apps

Whether you train a model from scratch or are able to use one that’s already trained, having access to a library of models in a standard interchange format will be a great productivity boost for Windows developers. Welcome to Machine Learning Studio, the Azure Machine Learning solution you have grown to love. Machine Learning Studio is a powerfully simple browser-based, visual drag-and-drop authoring environment where no coding is necessary. Go from idea to deployment in a matter of clicks.

If you are new to Machine Learning and have a strong wish to learn and master Azure ML and Win ML , this session is for you.

1 hour Deep Dive with Kubernetes

Micro-services and Docker has rapidly become the industry standard for containerizing applications, and are a terrific option for both greenfield .NET Core and legacy .NET Framework workloads.
In this session we will see how to leverage container technology to foster greater agility, efficiency, and security for enterprise organisations.
Come and see how to robustly move your applications to Docker containers which provides greater agility, efficiency, and the ability to migrate to modern infrastructure. Armed with this container knowledge In one hour we will be looking at many proven best practices and recipes to build rock-solid Kubernetes workloads.
You will leave this session with the knowledge and ability to design and develop your own Micro-services running on Kubernetes.

50 Shades of Containers

Benefit of running applications on Containers has been so evident, multiple teams within your organisation are now also interested in embracing Containers.
Expanding adoption of containers is exciting, however, are you ready to safely and efficiently manage containers in a your organisation.
In this talk, I will discuss about 50 recipes for robust and efficient container based design which will includes but not limited to container security, container scaling , optimising container cost , container monitoring, container logging, container storage and container availability.
As a result of attending this session you will be better prepared to consider and implement container based development approach and by the end of session you must also be aware about common challenges in container based development and how to resolve them efficiently.