By: TN Media News | Riyadh, KSA
A global survey commissioned by Ciena has revealed that AI workloads will be the primary driver of data center transformation over the next five years. The study, which included more than 1,300 data center decision-makers from 13 countries, forecasts at least a 6X increase in interconnect bandwidth demand during this period.
AI Workloads Lead Bandwidth Growth:
The research highlights that AI will surpass cloud computing and big data as the leading cause of data traffic growth. 43% of new data center builds are expected to be dedicated to AI workloads, given the immense data movement needed for training and inference of large AI models.
Fiber optic capacity is becoming a key consideration. A striking 87% of respondents said they will require 800 Gb/s or more per wavelength to keep up with growing demands.
Connectivity Now Equals Compute Power:
Jürgen Hatheier, CTO at Ciena, emphasized the urgent need for scalable infrastructure.
“AI is not just about compute; it’s about connectivity,” said Hatheier.
He added that data center operators must rethink network architectures to support accelerated traffic growth. This shift is crucial for enabling the real-time performance and efficiency that modern AI models demand.
Sustainability and Scalability Go Hand-in-Hand:
Sustainability was also flagged as a top concern. 98% of those surveyed agreed that using pluggable optics is essential to reduce power consumption and minimize network footprint.
Additionally, 81% foresee large language model (LLM) training being distributed across multiple data centers. This underscores the need for highly interconnected data center interconnect (DCI) networks.
Emerging Trends: Latency, Location, and Fiber:
The survey pointed to three major trends shaping the AI-ready infrastructure landscape:
-
63% prioritize AI resource utilization.
-
56% focus on reducing latency by placing compute closer to users.
-
54% are influenced by data sovereignty requirements.
Meanwhile, 67% of experts now prefer Managed Optical Fiber Networks (MOFN) for long-haul data transfers, signaling a shift from traditional dark fiber deployments.
Preparing for an AI-Driven Future:
As AI continues to evolve, data center operators must adapt quickly to the rising bandwidth and compute demands. The survey concludes that success in the AI era will depend on robust, flexible, and sustainable network infrastructure.
Ciena’s findings clearly show that staying ahead of the curve requires more than just processing power—it demands intelligent and scalable connectivity solutions.