AI expansion is urgent: companies are fighting against a lack of infrastructure!

AI expansion is urgent: companies are fighting against a lack of infrastructure!

The current study by Keysight Technologies entitled “Beyond the Bottleneck: AI Cluster Networking Report 2025” shows alarming trends in the development of artificial intelligence (AI) and the associated infrastructure. The results, developed in collaboration with Heavy Reading, reveal that the adoption of AI is progressing faster than the necessary infrastructure is available. These findings come from a comprehensive survey of telecom and cloud service provider professionals in various regions, including the Americas and Europe and Asia Pacific, between March and April 2025. The study highlights the urgent need for telecom and cloud service providers to move from simply expanding to optimizing their existing infrastructure.

A notable aspect of the survey is that 62% of respondents are looking for ways to get more out of their existing infrastructure without additional investment. Still, an overwhelming majority of 89% of survey respondents plan to expand or at least maintain their spending on AI infrastructure in the coming year. The reasons for this increase are varied: the strongest motivation comes from cloud integration (51%), followed by demand for faster GPUs (49%) and high-speed network upgrades (45%).

Challenges for AI development

Despite the high willingness to invest, the industry is faced with significant challenges. The biggest hurdle to scaling AI technologies is budget constraints, cited by 59% of respondents. Other limitations include infrastructural problems (55%) and a lack of qualified specialists (51%). These factors conflict with the goal of using and optimizing AI models efficiently. In particular, emulating real-world workloads is considered crucial to unlocking the full potential of existing infrastructure. 95% of respondents believe this is necessary, although many do not have the necessary tools.

In parallel to these developments, Microsoft is reporting a massive increase in its investments in AI infrastructure, which have increased by 79% compared to the previous year to now $14 billion. This increase exceeds company revenue growth of 17% and reflects growing demand for computing power and the integration of AI capabilities into products and services. Microsoft is also facing a lack of capacity, particularly in the inference phase of AI models, which has affected the company's AI results.

Strategic investments and outlook

However, through these challenges, Microsoft remains committed to significantly expanding its data center infrastructure. Cloud platform Azure reported 31% growth in the third quarter, of which 7 percentage points were directly attributable to AI integration. Amy Hood, Microsoft's chief financial officer, predicts a further increase in investments in the current quarter and advocates for higher spending in the following fiscal year to reflect increasing demands.

Interest in AI is growing across industries as companies increasingly recognize how AI can help improve their business processes. To remain a leader in AI, continued investment in AI infrastructure is critical for both Microsoft and many other companies. The focus on AI services within the Azure platform attracts new customers and therefore supports revenue growth, while lowering the barriers to using AI and strengthening the infrastructure for the future.

For the industry as a whole, the combination of this current data from the Keysight Technologies study and developments at Microsoft highlights the urgent need to invest in sustainable and powerful AI infrastructure to meet growing demand. This is the only way the industry can ensure that it remains at the cutting edge of technology while meeting the challenges that come with growing requirements and more complex AI models.

For more information about the study results, please visit pressagentur.com and on Microsoft's investments in AI infrastructure: finmagazin.de.