Connectivity Revolution or Evolution Inside Data Centers?

2026-06-08 · Source: Big Data & AI News - EE Times · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure, Emerging Technologies & Innovation · Depth: Advanced, medium

Summary

AI workloads are fundamentally transforming intra-data center connectivity, making networks as critical as compute. This shift blends revolutionary demands with optical evolution. Copper cables are being replaced by optical fiber for high-performance connections beyond 5 meters, with speeds advancing from 100 Gbps to 400 Gbps, 800 Gbps, and emerging 1.6 Tbps. This supports massive scale-out AI clusters, requiring 3-6 optical transceivers per GPU, totaling over a million short-reach units for hundreds of thousands of GPUs. LightCounting forecasts a doubling of Ethernet optical transceiver and co-packaged optics sales in five years. Current optical module designs (FRO, LRO, LPO) balance performance, power, and latency. Future innovations like Extra-dense Pluggable Optics (XPO), introduced in early 2026 for 12.8 Tbps, Near-Packaged Optics (NPO), and Co-Packaged Optics (CPO) aim to integrate optics closer to compute, enhancing efficiency and density.

Key takeaway

For AI Architects designing next-generation data centers, your network strategy must prioritize optical fiber and scale-out architectures. You should evaluate current options like LRO and LPO for efficiency in short-reach links, while actively planning for emerging technologies such as XPO, NPO, and CPO. Integrating optics closer to compute will be crucial for achieving the required density, power efficiency, and ultra-low latency demanded by future AI workloads.

Key insights

AI workloads necessitate a revolutionary shift to optical, scale-out intra-data center networks, integrating optics closer to compute.

Principles

• Intra-data center networks extend computing systems.
• Scale-out is critical for large AI model training.
• Optical module design prioritizes power and density.

In practice

• Transition to optical fiber for high-speed links.
• Utilize LRO/LPO for efficient short-reach connections.
• Explore XPO, NPO, CPO for future data center designs.

Topics

• Data Center Networking
• AI Accelerators
• Optical Transceivers
• Co-Packaged Optics
• Scale-Out Architecture
• Optical Fiber

Best for: Investor, CTO, VP of Engineering/Data, AI Architect, AI Hardware Engineer, MLOps Engineer

Related on AIssential

• Laser Chip Brings Multiplexing to AI Data Centers — Integrating DWDM lasers directly onto silicon wafers revolutionizes AI data center networking by enhancing bandwidth and reducing latency.
• Extreme Connect 2026: Momentum Depends On Platform ONE And AI — Extreme Networks is integrating AI into network operations to create business-optimized, autonomous network management.
• MWC 2026 Concludes as Telcos Pivot to AI — The telecom industry is debating GPU vs. CPU architectures for AI-driven 6G networks, impacting economics and deployment.
• 2026.23: Power Shifts — Elon Musk's companies transform market rules through scale and investor belief, exemplified by SpaceX's space data center vision.
• Taiwan Minister Emphasizes Collaboration and Future Focus on Photonics, WBG, and Quantum — Taiwan is strategically transitioning from a hardware manufacturer to a global AI innovation enabler through four key pillars.
• From building AI infrastructure to shaping its standards: Lambda joins OCP — AI infrastructure requires open, composable standards to keep pace with rapid hardware evolution and increasing density.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Big Data & AI News - EE Times.