FS Expands AI Optics Portfolio from 1.6T Scale-Out to 800G Scale-Across, Supporting Next-Gen AI Fabrics

September 13, 2026

Converge Digest Unknown

Overview

FS has significantly extended its optical connectivity portfolio for AI infrastructure, now offering 400G, 800G, and 1.6T optical transceivers for scale-out networks. Additionally, the company provides 400G and 800G coherent modules for scale-across data center interconnects, supporting distances up to 500km. This comprehensive portfolio is compatible with Ethernet, RoCE, and InfiniBand environments, enabling high-bandwidth, low-latency links between GPU servers, NICs, switches, and racks. The 1.6T generation, featuring 200G/lane electrical and optical interfaces, is positioned as crucial for next-generation AI fabrics.

In Depth

Key Findings

FS has announced a significant expansion of its optical connectivity portfolio specifically tailored for AI infrastructure, now offering a comprehensive range of solutions from 1.6T for scale-out networks to 800G for scale-across applications. This move aims to address the evolving and demanding requirements of modern AI computing environments.

Technical / Clinical Details

The expanded portfolio includes 400G, 800G, and 1.6T optical transceivers designed for scale-out networks within AI clusters, facilitating high-speed communication between GPU servers, Network Interface Cards (NICs), switches, and racks. For data center interconnects (DCI) that span geographical distances—referred to as “Scale-Across”—FS now offers 400G and 800G coherent modules capable of transmission distances up to 500km. This ensures robust connectivity for distributed AI infrastructures. The solutions support various network environments, including Ethernet, RoCE (RDMA over Converged Ethernet), and InfiniBand, which are critical for AI workloads requiring ultra-low latency and high bandwidth. Notably, the industry is transitioning to 200G/lane electrical and optical interfaces for 1.6T deployments, making this generation particularly vital for the performance of next-generation AI fabrics.

Background & Context

The proliferation of sophisticated AI models and large-scale machine learning requires an underlying infrastructure that can handle immense data traffic with unparalleled speed and efficiency. Traditional copper cabling and lower-speed optical modules are increasingly becoming bottlenecks, unable to meet the demands of tightly coupled GPU clusters. The shift towards higher-speed optical interconnects, especially at the 1.6T and beyond rates, is an industry imperative to enable the continued scaling of AI compute. FS’s strategic expansion directly responds to this market need, providing crucial components that enable data centers to build flexible, high-performance, and energy-efficient AI fabrics. The emphasis on coherent optics for longer-reach applications underscores the necessity of integrating geographically dispersed AI resources effectively.

Strategic Significance & Outlook

FS’s broadened AI optical connectivity portfolio provides crucial tools for AI infrastructure architects and operators, enabling them to maximize GPU cluster performance and build distributed AI systems. The focus on 1.6T and coherent technologies is integral to the next evolution of AI. This innovation will accelerate the realization of more efficient and higher-performing AI data centers, fostering the creation of new AI-driven services and applications. As a key partner in the AI ecosystem, FS is poised to continue its contributions at the forefront of optical communication technology, supporting the foundational requirements for advanced AI capabilities globally.

Source: https://convergedigest.com/fs-ai-optics-scale-out-scale-across-1-6t/

Get our weekly technology intelligence — free

Receive an infographic that lets you judge at a glance whether each field’s analysis report is worth reading.

Subscribe Free — Weekly Tech Intelligence

By subscribing, you’ll receive Troy-Technical’s weekly technology intelligence newsletter.