The landscape of edge computing is evolving, with 2025 emerging as a pivotal year that reshapes how stakeholders approach deployment, AI workloads, and infrastructure models, as indicated by research conducted by STL Partners.
Insights from the study reveal a notable shift in conversations surrounding edge computing. While the past discussions primarily revolved around low latency applications, the latest data demonstrates that data localization driven by regulatory and sovereignty considerations is now the primary driver for edge adoption, particularly in on-premise setups.
Challenges related to cost and complexity pose significant barriers, with nearly 60% of respondents identifying them as major obstacles hindering broader enterprise integration across various edge scenarios.
Regarding artificial intelligence (AI), the survey indicates that edge infrastructure is increasingly playing a crucial role in inferencing, while model training is anticipated to remain predominantly centralized. Sectors like defense and healthcare, which prioritize data security, are expected to explore federated learning. Moreover, a majority of experts predict that accelerated hardware such as GPUs will become standard in most edge deployments by 2030.
In terms of industry verticals, manufacturing continues to lead the pack in both on-premise edge adoption and its integration with operational technology (OT) systems.
The survey, conducted from July to September 2025, gathered insights from over 100 edge specialists from operators and software vendors, offering valuable perspectives on current sentiments and future trends. The comprehensive edge computing survey report is now accessible for further exploration.
Discover the latest trends and insights shaping the landscape of edge computing in 2025, with a focus on data sovereignty, AI advancements, and infrastructure models. Dive into the STL Partners research findings to gain valuable perspectives on the evolving edge deployment landscape.
Related
