Every year, Yotta brings together some of the sharpest minds in energy, AI, and networking to discuss the future of our digital ecosystem and the infrastructure underpinning it. The event is always alive with ideas, innovations, and predictions, but this year the conversation felt different.
From hyperscalers and data center operators to network engineers and AI architects, everyone in attendance shared the same sense of urgency – the systems we rely on today must evolve to meet the demands of tomorrow’s applications.
CEO at DE-CIX.
That snapshot revealed that our global networks are very much in a state of transition, shifting from best-effort to control, from consolidation to diversification, and even from the ground to the Earth’s orbit. Nearly half of those we asked believe enterprises will move away from the public Internet for mission-critical workloads.
A similar number expect the single-cloud era to end within the next year. Most see satellites becoming part of the digital backbone before 2030, and more than 60% anticipate connectivity being priced by latency rather than bandwidth.
All told, these findings highlight how quickly our assumptions about connectivity are changing.
Trust, performance, and proximity are emerging as the new foundations of digital infrastructure, signaling that the “new” Internet will be defined by more than just access and coverage – it will be defined by performance and intelligent interconnectivity.
The public internet – No longer the universal tool?
One of the most striking findings from our snap survey was the question of trust. When asked whether enterprises would still rely on the public Internet for their critical traffic by 2030, nearly half of respondents (48%) said no.
For decades, the Internet’s open and best-effort model served as the foundation for global innovation, collaboration, and commerce. But that same model has become a growing liability for organizations operating in an environment where performance SLAs, resilience, data sovereignty, and compliance are now critical.
Outages, cyberattacks, and unpredictable data traffic routing have eroded confidence in the public Internet, particularly among enterprises with distributed workloads and working in critical sectors with territorial regulatory obligations.
The rise of zero-trust architectures and tightening security frameworks have only added to this demand for more controlled forms of connectivity – in other words, networks where data sovereignty, performance, and accountability can be clearly defined and the results measured.
Multi-cloud becomes the status quo
Another revelation was that we might well be approaching the end of the single-cloud era. Almost half of the industry experts we approached (48%) expect enterprises to abandon single-cloud strategies altogether in the next 12 months, giving new meaning to the concept of “cloud-first” innovation.
While early cloud migrations prioritized convenience and scale, dependency started to deepen and risks like outages, price volatility, and regional compliance penalties started to mount up.
Now, agility and choice are the lead priorities, with enterprises distributing workloads across multiple cloud providers, combining public, private, and edge environments to balance cost, performance, resilience, and sovereignty.
The driving force behind this diversification used to be avoiding vendor lock-in, but now it’s an architectural principle – “agility through diversification.”
However, this path does come with its own set of challenges. Managing data and application flows across disparate clouds, each with their own interfaces, compliance obligations, and performance profiles, requires a connectivity ecosystem capable of stitching them together seamlessly.
AI is adding to this complexity, where training may occur in one environment, inference in another, and the data fueling both may reside somewhere entirely different. Everything is becoming interdependent, and that makes connectivity the new strategic control plane.
Latency as the new currency, perhaps literally
If one theme united almost every discussion at Yotta 2025, it was latency. Almost two-thirds (61%) believe that within the next decade, connectivity will be priced not by bandwidth but by latency.
For decades, speed has been quantified in megabits and gigabits per second, but that’s a reflection of capacity rather than responsiveness. The continued growth of AI inference, autonomous systems, industrial automation, and telemedicine has left zero tolerance for delay.
In these applications, milliseconds can mean the difference between a system working as it should or verging on completely unusable. There is no middle ground.
So, a new logic is emerging for network and data centers. Operators are redesigning interconnection capabilities for ultra-low latency routing; data centers are being built closer to the edge, and enterprises are starting to buy “proximity-as-a-service.”
Latency has therefore become a unit of value – the faster data can circulate through a distributed system, the smoother the operation and the better the financial return. As I see it, the next decade will transform latency from a simple technical metric into a business building one.
The sky becomes part of the network – satellites enter the backbone
When asked whether satellite Internet would become as common as fiber by 2030, 65% of our respondents said yes. More than half went further, predicting that the first Internet Exchange (IX) could be operating in orbit within the decade.
A few years ago, this may have sounded like science fiction, but today, it reflects a shift toward the kind of “big thinking” that will be needed to support and sustain a growing number of latency-sensitive, AI-based use cases.
Low Earth orbit (LEO) constellations have done well to transform the perception of satellite connectivity and establish it as part of our digital backbone, but latency has been a persistent issue.
With optical and laser-based transmission systems such as those being explored in the European Space Agency’s OFELIAS project, the connectivity between satellites and terrestrial facilities will not only be fast, but also more resilient and stable.
Our pulse survey at Yotta 2025 may not have been scientific in its methodology, but it did give us a solid insight into how the world’s networks will connect in the coming decade. Networks will become more widely distributed and increasingly performance driven, while routing will become more deliberate.
As we extend connectivity across clouds, continents, and even in space, our biggest challenge will be to interconnect intelligently, ensuring that every millisecond and every link strengthens the connective fabric of the digital economy.
We've featured the best database software.
This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.