I’ve to confess, heading out to London for 2025 KubeCon + CloudNativeCon Europe, I assumed I’d see the start of the downward pattern for the occasion about constructing, deploying, and managing next-generation cloud purposes and infrastructures.
In spite of everything, the present turned 10 final yr, and, in my expertise, that is when conferences begin to present their age. Plus, there was numerous information round the impact of AI on utility growth, and whereas KubeCon is not straight about dev, a lot of its focus is on purposes and companies.
However boy, was I unsuitable. In actual fact, KubeCon 2025 in London was packed, with over 12,000 attendees. Getting by the crowds on the large London ExCel middle at occasions made me really feel like an contaminated Brad Pitt attempting to chop by an oncoming horde of zombies in World Warfare Z. And good luck moving into any of the keynotes even just a little bit late, with each seat in the principle corridor and overflow balconies taken.
So, what’s behind the keenness for core infrastructure and cloud know-how? One key to success could also be that Kubernetes and different core cloud-native applied sciences are proving pretty AI-resistant, particularly in contrast with common utility coding.
The training and data from a KubeCon occasion are centered on deployment, administration, and integration points that require experience and an understanding of complicated applied sciences. Plus, the fast tempo of change in cloud native might be robust for even latest fashions to maintain up with, and anybody utilizing AI will in all probability be constructing very out-of-date techniques.
On prime of those considerations, Kubernetes and cloud-native applied sciences are key to working and constructing AI workloads, with many prime techniques based mostly on cloud-native infrastructures. In actual fact, OpenAI was constructing its techniques on Kubernetes again in 2016.
Like final yr’s occasion, a lot of the dialogue round AI was eminently sensible and centered on key areas like automation and analytics, that are additionally prime use circumstances recognized by Aberdeen analysis.
One other fascinating pattern at KubeCon London was the return of the non-public cloud. Now, non-public cloud by no means went away however its significance had positively waned, with one main know-how vendor telling me a few years in the past that they weren’t even mentioning non-public cloud to purchasers anymore, with most of their messaging centered on hybrid cloud.
Nonetheless, in lots of my conferences and demos at KubeCon London, discussions round non-public cloud had been entrance and middle. Two main developments are driving this resurgence.
The primary pattern, as with many issues in know-how, is the rise of AI. In our analysis at Aberdeen, we discover that many companies are working to construct small language fashions or do inside inferencing to make the most of their information in AI.
When companies take this strategy, they rapidly understand they have to shield and safe this important non-public information, which requires avoiding public cloud AI and constructing AI infrastructures utilizing non-public cloud.
The opposite large pattern driving the return of personal cloud, as was talked about in lots of classes and discussions at KubeCon Europe, was the significance of digital sovereignty. With world uncertainty on all fronts, and plenty of companies involved about who may demand entry to their information in clouds across the globe, I heard from many distributors and companies seeking to deploy modernized non-public cloud infrastructures to make sure information integrity and hold their data protected.
This pattern reveals the elevated significance of the non-public cloud and the maturity of newer implementations of this provision. Whereas the non-public cloud began as primarily a strategy to carry cloud-based controls to on-prem techniques, it now provides companies all the agility and adaptability of the cloud with a lot better management and safety.
As companies look to modernize their infrastructure, shield their information, and leverage applied sciences like AI, Kubernetes and cloud-native applied sciences will most probably keep entrance and middle. It is going to be fascinating to see at this fall’s occasion in Atlanta if KubeCon will proceed to develop and keep important.
Need extra tales about AI? Join Innovation, our weekly e-newsletter.
