I’m Speaking at Red Hat Summit 2026!
I’m thrilled to announce that my proposal has been accepted for Red Hat Summit 2026 in Atlanta, Georgia! Out of a highly competitive pool of submissions, my lightning talk on GPU multi-tenancy and platform engineering was selected — and I couldn’t be more excited to share our production experience with the Red Hat community.
Watch the Video
Session Details
| Detail | Information |
|---|---|
| Event | Red Hat Summit 2026 |
| Title | GPUs take flight: Safety-first multi-tenant Platform Engineering with NVIDIA and Red Hat OpenShift AI |
| Session Type | Lightning Talk (20 min) |
| Date | Tuesday, May 12, 2026 |
| Time | 10:30 AM EDT |
| Room | Discovery Theater 1 |
| Location | Atlanta, Georgia, USA |
| Speaker | Luca Berton |
Talk: GPUs Take Flight — Safety-First Multi-Tenant Platform Engineering with NVIDIA and Red Hat OpenShift AI
Abstract
As organisations race to deploy AI/ML workloads, GPU infrastructure is becoming the most expensive and contested resource in enterprise IT. This lightning talk distils real-world lessons from building a safety-first, multi-tenant GPU platform using NVIDIA hardware and Red Hat OpenShift AI.
What You’ll Learn
- Multi-Tenant GPU Isolation: How to share GPU resources across teams safely, without noisy-neighbour problems
- Platform Engineering Patterns: Building self-service GPU platforms that data scientists actually want to use
- NVIDIA + OpenShift AI Integration: Practical guidance on combining NVIDIA GPUs with Red Hat’s AI/ML platform
- Safety-First Architecture: Guardrails, quotas, and policies that prevent runaway workloads from burning through your GPU budget
- Production Lessons: Battle-tested patterns and pitfalls from running multi-tenant GPU workloads at scale
Why This Talk Matters
GPU compute is the engine behind the AI revolution — but sharing it safely across multiple teams and workloads remains one of the hardest unsolved challenges in platform engineering.
For Platform Engineers
- Build GPU platforms with confidence: Learn isolation patterns that go beyond basic namespace quotas
- Self-service done right: Enable data science teams to provision GPU workloads without compromising cluster stability
- Day-2 operations: Observability, chargeback, and lifecycle management for GPU infrastructure
For AI/ML Teams
- Maximise GPU utilisation: Move from single-tenant waste to efficient multi-tenant sharing
- Faster time to production: Platform engineering patterns that reduce the gap between experimentation and deployment
- Cost transparency: Understand exactly what your GPU workloads cost and how to optimise spend
For Decision Makers
- Enterprise-grade GPU strategy: How Red Hat OpenShift AI and NVIDIA create a production-ready foundation
- Risk mitigation: Safety-first architecture patterns that protect your investment and your SLAs
This isn’t a vendor pitch — it’s an experience report from the trenches of GPU platform engineering. Every recommendation comes from real production workloads.
Register for Red Hat Summit 2026
Red Hat Summit is the premier enterprise open-source technology event, bringing together thousands of IT professionals, developers, and business leaders.
👉 Register at summit.redhat.com
🎤 View my session in the catalog →
What Is Red Hat Summit?
Red Hat Summit is Red Hat’s flagship annual conference — the premier event for enterprise open-source innovation. It brings together customers, partners, community contributors, and Red Hat engineers to explore the latest in hybrid cloud, AI/ML, automation, and security.
By the Numbers
| Metric | Scale |
|---|---|
| Attendees | 7,000+ in-person + virtual |
| Sessions | Breakout sessions, lightning talks, labs, panels, and product roadmaps |
| Duration | 4 days (May 11–14, 2026) |
| Location | Atlanta, Georgia, USA |
| Highlights | Keynotes, Expo Hall, Executive Exchange, Labs, Georgia Aquarium closing event |
Why Red Hat Summit Matters
Red Hat Summit is where the future of enterprise open source is shaped. It’s where:
- Product roadmaps are revealed and discussed with the engineers who build them
- OpenShift, RHEL, Ansible, and AI/ML innovations are previewed first
- Enterprise adoption patterns are shared through real-world customer case studies
- The AI on OpenShift movement is accelerating — GPU orchestration, model serving, and MLOps at enterprise scale
- Partner ecosystem solutions from NVIDIA, Dell, Intel, AWS, and more are showcased
Event Agenda Highlights
| Day | Date | Highlights |
|---|---|---|
| Day 1 | Monday, May 11 | Community Day, Ecosystem Success Day, Kickoff Party |
| Day 2 | Tuesday, May 12 | Keynote, Sessions, Labs, Expo Hall, Expo Networking Reception |
| Day 3 | Wednesday, May 13 | Keynote, Sessions, Labs, Expo Hall, Georgia Aquarium Closing Event |
| Day 4 | Thursday, May 14 | Sessions, Labs, Late-Night Labs |
Stay Updated
Want to follow my Red Hat Summit journey? Subscribe and follow for updates:
Thank You
A huge thank you to the Red Hat Summit content team for selecting this talk! I’m looking forward to sharing production-tested patterns for GPU multi-tenancy with the Red Hat community.
See you in Atlanta! 🇺🇸
Related
Have questions about GPU platform engineering or OpenShift AI? Drop a comment or reach out on social media — I’m always happy to chat about multi-tenant GPU infrastructure!
