SINGAPORE, September 5, 2025 – Artificial intelligence is advancing faster than the infrastructure designed to support it, according to a new global study released by Keysight Technologies in collaboration with Heavy Reading.
The Beyond the Bottleneck: AI Cluster Networking Report 2025 found that while nearly nine in ten telecom and cloud providers plan to maintain or expand AI-related investments over the next year, many face growing constraints in scaling their data center operations.
The report highlights an industry-wide pivot: building out infrastructure is no longer sufficient. Sixty-two percent of respondents said their priority is extracting more value from existing systems without new capital expenditures.
Operators are increasingly adopting performance optimization strategies such as real-world AI workload emulation, which allows them to validate performance, improve efficiency, and accelerate deployment of next-generation AI clusters.
Key Findings
Optimization First: 62% of operators are prioritizing efficiency over expansion to address cost pressures.
Emulation as Standard: 95% of respondents view real-world workload emulation as critical, though many still lack tools to simulate production-scale AI environments.
Scaling Barriers: Budget constraints (59%), infrastructure limitations (55%), and talent shortages (51%) remain the biggest hurdles.
High-Speed Networking: 34% are exploring 800G, 22% trialing 1.6T, and 58% evaluating Ultra Ethernet for high-performance networking.
Bottlenecks Emerging: With 55% of operators deploying 400G interconnects and increasing trials of 1.6T, network capacity is becoming a decisive factor for AI scalability.
Industry at a Tipping Point
“AI data centers are reaching a tipping point where performance and scale alone are not enough. Operators need deeper insight, tighter validation, and smarter infrastructure choices,” said Ram Periakaruppan, Vice President and General Manager, Network Applications & Security Group at Keysight.
“This research confirms what we see in the field: success in the AI era hinges on optimizing every layer of the network. Keysight is proud to support this shift with solutions that enable providers to emulate, validate, and future-proof their AI infrastructure at scale.”
Broader Implications
The findings underscore a fundamental shift in how telecom and cloud service providers are preparing for the AI-driven future. With cloud integration (51%), faster GPUs (49%), and high-speed network upgrades (45%) cited as top drivers of investment, the focus is moving beyond raw capacity to performance and reliability.
As AI models continue to grow more complex, industry leaders say the next phase of competition will be defined by how efficiently infrastructure can be optimized to deliver scalable, cost-effective, and reliable outcomes.