Private AI Sessions at Explore Barcelona

Attention, AI enthusiasts! VMware Explore is just around the corner, and the buzz about artificial intelligence is stronger than ever. With a staggering 148 sessions mentioning AI, it's clear that private AI is taking center stage. To help you navigate this wealth of knowledge, I've curated a day-by-day list of must-attend sessions focused on private AI, specifically at the intermediate (200) and advanced (300) levels. Whether you want to deepen your expertise or explore cutting-edge concepts, this guide will ensure you don't miss out on the most impactful insights and innovations. Get ready to fill your schedule with private AI tailored for those ready to dive deep!

Monday 4 November

Empowering Developers with Open Source GenAI Models [INVB1570BCN]

10:15 - 11:00 CET - Ramesh Radhakrishnan - Level 200

Session Summary:

1. VMware Cloud Foundation's team has implemented a GenAI stack for open-source models, enabling developers to leverage AI capabilities.
2. The presentation will discuss challenges in hosting advanced models and innovative solutions.
3. Key highlights include: Open Source Approach: Benefits of using open-source technologies Scalable Infrastructure: Architecture insights, including distributed computing and resource management on VMware Cloud Foundation Developer Experience: User-friendly APIs and resources for effective GenAI model utilization Model Serving: Streamlined processes and tools for a seamless user experience
4. The focus is empowering developers and teams to use GenAI models within the VMware Cloud Foundation ecosystem.

Accelerate AI Implementation and Improve TCO with Intel Xeon [VCFT1831BCNS]

11:30 - 13:00 CET - Simon Todd MSc & Patryk Wolsza - Level 200

Session Summary:

1. AI integration with Intel AMX on VMware Cloud Foundation offers businesses enhanced operational efficiency, scalability, and competitiveness.
2. The session explores the synergy between data center modernization and AI, focusing on performance, security, and agility improvements.
3. Attendees will gain insights into strategic implications, implementation best practices, and real-world use cases.
4. The presentation aims to redefine data center innovation and AI boundaries.

Building GenAI Apps on VMware Cloud Foundation [INVB1561BCN]

12:45 - 13:30 CET - Ramesh Radhakrishnan & Ala Solsvig Dewberry - Level 200

Session Summary:

1. Organizations with untapped data potential can gain a competitive advantage using AI capabilities like Retrieval-Augmented Generation (RAG).
2. Building AI-powered chat applications presents challenges in model management, data privacy, and infrastructure enablement.
3. VMware Private AI Foundation on VMware Cloud Foundation offers a streamlined solution for developing and deploying AI-powered chat applications with enterprise risk management.
4. The presentation will teach attendees how to: Create secure resource envelopes for GenAI workloads. Set storage policies and manage data access. Securely manage models, data, and chat engines. Build a GenAI app using a user-friendly CLI, API, or UI.

Deep Dive: Running VMware Private AI Foundation with NVIDIA on VCF [VCFT1804BCN]

14:00 - 15:30 CET - Justin Murray - Level 300

Session Summary:

1. VI Admins can extend their skillset with new AI capabilities offered by VMware Cloud Foundation, enabling secure support for business needs in the evolving AI landscape.
2. The platform supports running AI/ML workloads with integrated NVIDIA technology and allows automated provisioning of AI-ready infrastructure using Kubernetes and the IaaS control plane.
3. VMware Cloud Foundation facilitates the deployment of generative AI large language models (LLMs) and streamlines the development and operations of AI applications.
4. Data Services Manager (DSM) plays a crucial role in supporting AI workloads by providing essential data services, offering vector database capabilities, and enabling seamless integration with VMware Private AI Foundation.

Tuesday 5 November

Keynote 09:00 - 10:00

Building Service Deliverables with VMware Private AI Foundation with NVIDIA [CODEB1186BCN]

10:30 - 11:15 CET - Frank Denneman & Johan van Amersfoort - Level 300

Session Summary:

1. ITQ, with extensive enterprise IT experience, faces challenges in maintaining consistent service delivery across numerous engagements and customers as their team grows.
2. The company implemented a service-oriented retrieval-augmented generation (RAG) application using VMware Private AI Foundation with NVIDIA to streamline knowledge management and decision-making processes.
3. The presentation will explore the technical aspects of the Private AI Foundation platform, including its design, integration with VMware Cloud Foundation, hardware considerations, and NVIDIA NIM deployment.
4. Attendees will see demonstrations of the virtual ITQ assistant, showcasing how it improves knowledge aggregation, sales offer creation, and helpdesk support.

Building cost-effective home labs for AI/ML workloads [CMTY1447BCN]

14:00 - 14:25 CET - Johan van Amersfoort - Level 200

Session Summary:

1. AI workloads can be run in a home lab environment without investing in expensive data center-grade GPUs.
2. The session covers Johan's 7-year journey of building and experimenting with AI workloads in his home lab, sharing insights on what worked and what didn't.
3. Johan demonstrates that it's possible to work with AI and machine learning without always needing a GPU, making AI experimentation more accessible2.
4. The presentation includes practical examples and demos, such as building an AI detector for identifying when a neighbor's cat is in your yard and showcasing the potential of home-based AI projects.

Wednesday 6 November

Accelerate Intelligent App Delivery with VMware Tanzu AI and VMware Private AI Foundation [TANB1617BCN]

09:00 - 09:45 CET - Adib Saikali & Sesh Sayani - Level 200

Session Summary:

1. VMware Tanzu AI enables developers to create intelligent AI-powered apps without requiring extensive Python or LLM expertise.
2. The integration of VMware Tanzu AI, VMware Private AI Foundation, and NVIDIA allows for the secure development, operation, and optimization of AI-powered applications.
3. This solution addresses governance and security challenges platform engineers and security leaders face when utilizing large language models.
4. The session covers VMware Tanzu AI fundamentals and demonstrates how to accelerate app delivery using VMware Tanzu Platform and VMware Private AI Foundation with NVIDIA, focusing on consumable AI frameworks.

Unlocking Your Data with VMware and NVIDIA—RAG Deep Dive [VCFB1881BCN]

10:15 - 11:00 CET - Frank Denneman & Justin Murray - Level 300

Session Summary:

1. Retrieval Augmented Generation (RAG) combines large language models with current, relevant data from trusted sources to address the challenge of accessing the right information from massive datasets.
2. The session provides a 300-level examination of a RAG pipeline architected using Private AI Foundation with NVIDIA, including an overview of the end-to-end workflow and its mapping to virtual infrastructure.
3. The presentation delves into the components of the RAG pipeline, such as the Large Language Model (LLM), NVIDIA Inference Microservices, and Vector Database.
4. Best practices for implementing RAG are discussed, including sizing considerations and guidance on when to use fractional, full, or multiple GPUs for optimal performance.

Private AI and Advanced Services on VCF: What's Next? [INVB1931BCN]

11:30 - 12:15 CET - Chris Wolf & Dilpreet Bindra - Level 300

Session Summary:

1. VMware Cloud Foundation will leverage AI to streamline product support and troubleshooting processes, enhancing overall platform efficiency.
2. The platform will provide improved capabilities for securely and efficiently operating AI applications and models.
3. New features will be introduced to enable running a variety of Platform-as-a-Service (PaaS) and data services on VMware Cloud Foundation.
4. The presentation will showcase first-ever demos of emerging capabilities and highlight collaborative efforts with ecosystem partners and customers to develop new solutions.

Build your first AI Proof of Concept on VCF! [CMTY1928BCN]

14:00 - 14:25 CET - Robert Kloosterhuis & Johan van Amersfoort - Level 200

Session Summary:

1. The session addresses the challenge of supporting AI workloads for data science teams, targeting those who need help determining where to begin.
2. Johan van Amersfoort and Robert Kloosterhuis will guide attendees through building a VMware Private AI Foundation proof of concept.
3. The presentation includes setting up the IaaS Control Plane as part of the VMware Private AI Foundation implementation.
4. The speakers will share lessons learned from their experience, offering attendees a fun and interactive session.

AI Without GPUs: Using Your Existing CPU Resources to Run AI Workloads [VCFB1882BCN]

15:15 - 16:00 CET - Will Rodbard & Frank Denneman - Level 300

Session Summary:

1. Many AI/ML applications can run effectively on CPUs despite GPUs being powerful but expensive and difficult to acquire.
2. The session explores leveraging existing CPU resources for AI workloads, mainly focusing on Intel Xeon 4th Gen CPUs with Advanced Matrix Extensions (AMX).
3. Attendees will learn about the capabilities and limitations of CPU-based AI processing, including what types of problems can be solved without GPUs.
4. The presentation includes demonstrations of CPU-optimized AI applications, such as computer vision using OpenVINO and large language models (LLMs) tuned for AMX.

Enjoy your week at Explore! If you see me walk by, don't hesitate to reach out and have a chat about your AI experience and challenges.