I’m Speaking at KubeCon Europe 2026!
I’m thrilled to announce that I’ll be presenting at KubeCon + CloudNativeCon Europe 2026 in Amsterdam! This is the world’s largest and most important Kubernetes conference, and being selected as a speaker is an incredible honour. Out of thousands of submissions, only a fraction make the cut — and I’m proud to bring real-world production experience to the stage.
Session Details
| Detail | Information |
|---|---|
| Event | KubeCon + CloudNativeCon Europe 2026 |
| Title | Lessons Learned Orchestrating Multi-Tenant GPUs on OpenShift AI with NVIDIA KAI (G/H200) |
| Date | Tuesday, March 24, 2026 |
| Time | 16:15 – 16:45 CET |
| Location | RAI Amsterdam — Hall 8, Room F |
| Level | Intermediate |
| Track | AI + ML |
| Affiliation | Dell Technologies |
Talk: Lessons Learned Orchestrating Multi-Tenant GPUs on OpenShift AI with NVIDIA KAI (G/H200)
Abstract
How do you run shared, production-grade GPUs for AI/ML safely and efficiently? This experience report distils hard-won lessons from implementing multi-tenant GPU orchestration on OpenShift AI using NVIDIA KAI on G/H200 hardware, fronted by Traefik and backed by Dell Technologies platforms.
What You’ll Learn
- Tenant Isolation Patterns: Namespaces, quotas, and priority classes for safe multi-tenancy
- Heterogeneous Node Scheduling: Placing workloads intelligently across mixed GPU types
- MIG vs. Full-GPU Trade-offs: When to partition and when to dedicate entire GPUs
- Throughput vs. Latency Tuning: Optimising for batch training vs. real-time inference
- Driver & Firmware Pitfalls: Lessons from the trenches on GPU driver lifecycle
- Upgrade & Rollback Strategies: Zero-downtime GPU stack updates
- Day-2 Operations: Observability, autoscaling, and chargeback for GPU compute
- Practical Manifests & Guardrails: Ready-to-use configurations you can apply immediately
Why This Talk Matters for the Ecosystem
GPU compute is the most expensive and contested resource in modern IT. As every organisation races to deploy AI/ML workloads, the Kubernetes ecosystem faces a critical challenge: how do you share GPU infrastructure safely, efficiently, and at scale?
This talk matters because:
For the Cloud-Native Community
- GPU multi-tenancy is largely uncharted territory — most Kubernetes documentation and tooling assume CPU-based workloads. This talk fills a real knowledge gap with production-tested patterns.
- NVIDIA KAI on Kubernetes is still new — attendees get early, honest insights into what works and what breaks.
- OpenShift AI + Traefik + Dell is a real enterprise stack — not a demo on a single node, but lessons from running at scale.
For AI/ML Teams
- Reduce GPU waste: Learn isolation and scheduling patterns that push utilisation from 30% to 80%+.
- Avoid costly mistakes: Driver/firmware issues can take down entire GPU pools — hear how to prevent them.
- Budget transparency: Chargeback and observability patterns so teams know exactly what GPU time costs.
For Platform Engineers
- Battle-tested manifests: Walk away with YAML you can deploy on Monday morning.
- Upgrade without downtime: GPU driver and firmware lifecycle strategies that don’t require maintenance windows.
- Autoscaling GPU nodes: Patterns for scaling expensive hardware responsibly.
This isn’t a theoretical talk — it’s an experience report from running production GPU workloads. Every recommendation comes from something that either worked brilliantly or failed spectacularly.
Join Me at KubeCon!
If you’re planning to attend KubeCon Europe 2026, I’d love to see you there!
What Is KubeCon + CloudNativeCon?
KubeCon + CloudNativeCon is the flagship conference of the Cloud Native Computing Foundation (CNCF) — the vendor-neutral home of Kubernetes, Prometheus, Envoy, and 180+ other open-source projects that power modern infrastructure.
It is, quite simply, the most important event in cloud-native computing.
By the Numbers
| Metric | Scale |
|---|---|
| Attendees | 12,000+ in-person + thousands virtual |
| Sessions | 300+ talks, workshops, and tutorials |
| Sponsors | 200+ companies from startups to hyperscalers |
| Co-located Events | 20+ including Istio Day, ArgoCon, Platform Engineering Day |
| Countries Represented | 100+ |
Why KubeCon Matters
KubeCon is where the future of infrastructure is decided. It’s where:
- Kubernetes releases are previewed and discussed by the people who build them
- New CNCF projects are announced (graduated, incubating, sandbox)
- Enterprise adoption patterns are shared through real-world case studies
- The AI/ML on Kubernetes movement is accelerating — GPU scheduling, model serving, MLOps pipelines
- Platform engineering best practices are shaped by practitioners, not vendors
Being selected to speak at KubeCon means your work has been peer-reviewed by the program committee and deemed valuable to the global community. It’s the gold standard for cloud-native thought leadership.
Why Attend?
- Technical Deep-Dives: From Kubernetes internals to AI/ML orchestration
- Networking: Meet the maintainers, contributors, and adopters shaping the ecosystem
- Sponsor Showcase: Explore the latest tools and platforms from 200+ companies
- Hallway Track: Some of the best conversations happen between sessions
- Career Growth: KubeCon speakers and attendees are at the forefront of the industry
Stay Updated
Want to follow my KubeCon journey? Subscribe and follow for updates:
Thank You
A huge thank you to the KubeCon program committee for selecting this talk. I’m looking forward to sharing our experiences and learning from the incredible cloud-native community!
See you in Amsterdam! 🇳🇱
Related
Have questions about GPU orchestration or OpenShift AI? Drop a comment or reach out on social media — I’m always happy to chat about cloud-native infrastructure!
