GIGALIGHT Unveils 18W Ultra-Low Power 1.6T AI Interconnect at ECOC 2026, Signaling Independence from Liquid Cooling

September 20, 2026

GIGALIGHT Press Releases China

Overview

GIGALIGHT showcased a hybrid technology at ECOC 2026 for 1.6T AI interconnects, achieving a significantly reduced power consumption of 18W compared to mainstream 3nm full-DSP solutions. This innovation aims to alleviate thermal pressure in high-density OSFP switch racks, enabling AI training clusters to scale without sole reliance on liquid cooling. The solution provides robust transmission and BER performance exceeding international protocol specifications, supporting 500-meter short- and medium-distance interconnects in AI clusters.

In Depth

Key Findings

At ECOC 2026, GIGALIGHT unveiled a 1.6T AI interconnect featuring a hybrid technology that drastically reduces power consumption to 18W, a significant improvement over mainstream 3nm full-DSP solutions. This breakthrough aims to mitigate thermal challenges in high-density OSFP switch racks, allowing AI training clusters to scale without exclusive reliance on liquid cooling systems.

Technical / Clinical Details

GIGALIGHT’s presented 1.6T AI interconnect utilizes an innovative hybrid technology to achieve remarkable power efficiency. Specifically, it reduces power consumption to a mere 18W, a substantial decrease from the levels typically required by current leading 3nm process technology-based full-DSP solutions. This significant power reduction substantially alleviates thermal design challenges in AI data centers, particularly in high-density OSFP switch racks populated with numerous GPUs and accelerators. Lowering the thermal load simplifies rack cooling requirements and directly translates into reduced operational costs. The solution delivers robust transmission performance and Bit Error Rate (BER) exceeding international protocol specifications, ensuring stable operation for short- to medium-distance interconnects up to 500 meters within AI clusters. This enables high-speed, reliable data transfer for AI workloads and facilitates flexible deployment and scaling of AI clusters even in environments where liquid cooling is not mandatory.

Background & Context

The increasing demands of AI computing are introducing new challenges for power consumption and thermal management in data centers. Particularly in AI training clusters, where a large number of GPUs operate simultaneously, the generated heat can reach levels that threaten system performance and reliability. While liquid cooling systems offer an effective solution, they often come with high implementation costs and increase infrastructure complexity. GIGALIGHT’s hybrid technology addresses this issue by offering a different approach: fundamentally reducing power consumption itself. This represents a crucial step towards building more sustainable and cost-effective AI infrastructure, drawing significant industry attention.

Strategic Significance & Outlook

GIGALIGHT’s ultra-low power 1.6T AI interconnect technology holds the potential to profoundly impact the design and operation of AI data centers. The ability to reduce reliance on liquid cooling will accelerate the deployment of AI clusters, especially in environments with space constraints or limited budgets. As this technology gains wider adoption, it is expected to enhance the overall energy efficiency of AI infrastructure and facilitate the creation of larger, more distributed AI computing environments. GIGALIGHT aims to leverage this technology to establish itself as a key player in the next-generation AI data center market.

Source: https://www.gigalight.com/news-events/news-10875.html

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.