Keysight Study Reveals AI is Outpacing Infrastructure Readiness

Keysight Technologies has released a new report in collaboration with Heavy Reading, exploring how operators are evolving to meet the scale and speed of AI workloads. The new study, titled, Beyond the Bottleneck: AI Cluster Networking Report 2025, revealed that nearly 9 in 10 (89%) telecom and cloud providers plan to expand or maintain AI infrastructure investments, with the top three growth factors cited as cloud integration (51%), faster GPUs (49%), and high-speed network upgrades (45%).

As artificial intelligence accelerates across industries, the demands on data center infrastructure are rising fast. Building out infrastructure is no longer enough on its own; 62% of respondents said they aim to get more out of their infrastructure without new investment. Instead, operators are turning to performance optimization strategies, such as real-world AI workload emulation, not only to validate performance but also to enhance efficiency and accelerate the deployment of next-gen AI clusters.

Despite this momentum, over 50% of operators cite budget constraints (59%), infrastructure limitations (55%), and talent shortages (51%) as their top three roadblocks. At the same time, the need for high-speed Ethernet networking is rising sharply, with strong interest in 800G (34%), 1.6T (22%), and Ultra Ethernet technologies (58%). As a result, providers are shifting focus from expansion to optimization, aiming to boost performance, efficiency, and reliability without additional spending. 

“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.”

Other key findings from the Beyond the Bottleneck: AI Cluster Networking Report 2025 include: 

• Emulation Is Now Essential: 95% of respondents said real-world workload emulation is critical, but most lack the tools to effectively mirror production-scale AI environments
• Network Bottlenecks Are the New Frontier: 55% of operators are deploying 400G interconnects, while 22% are already trialing 1.6T, making network capacity a decisive factor in AI scalability
• Ultra Ethernet Gains Momentum: 58% are evaluating Ultra Ethernet as a high-performance networking option.

The study demonstrates that the evolution of AI infrastructure is as much about optimization as it is about capacity. As operators adapt to the demands of increasingly sophisticated AI models, tools and techniques that maximize efficiency and performance, such as real-world AI workload emulation, are crucial for maximizing the potential of data center infrastructure. This transformation reflects an industry shift towards efficiency, reliability, and adaptability. To stay ahead, operators must combine smart investments with smarter operations. 

Keysight offers a portfolio of integrated design, test, and validation solutions, purpose-built for high-performance, AI-optimized data center environments. 

About the Survey:

The data in the AI Cluster Networking Report was based on a global survey conducted globally by Heavy Reading between March and April 2025. The study collected insights from professionals across leading service providers from the Americas, Asia Pacific, and Europe.

Beyond the Bottleneck: AI Cluster Networking Report 2025 explores the needs of network operators as they evolve to meet the scale and speed of AI workloads. Keysight collaborated with Heavy Reading on the report.