Chroma — leaderboard

Edgecore launches 102.4T data center switches for 800G and 1.6T AI fabrics

Edgecore Networks has announced two 102.4T open-networking data center switches, the AIS1600-64O and AIS800-128O, aimed at hyperscale artificial intelligence and machine learning clusters and next-generation cloud fabrics. The company says both platforms are powered by Broadcom Tomahawk 6 silicon built on a 3 nm process, targeting high bandwidth and low latency for large-scale fabrics.

The portfolio includes two form factors and port profiles. The AIS1600-64O is a 3RU system with 64 OSFP1600 ports at 1.6 T per port, positioned for environments moving to 1.6T Ethernet. Edgecore says it supports Flexport technology for “breakout configurations from 100G to 1.6T” and is intended to support “upcoming 800G NIC generations.” The AIS800-128O is a 4RU switch with 128 OSFP800 ports at 800G; Edgecore says it can expose “up to 512 logical ports on a single chip” to increase fabric connectivity density, reduce hop count, and lower network latency, including for “the largest 2-tier scale of 400G NIC connectivity” using 800G breakout.

On fabric behavior and resiliency, Edgecore says the AIS series reduces job completion time through “hardware-based link failover and intelligent load balancing (DLB/GLB),” and is designed to prevent bottlenecks in “high-scale AI fabrics.” For power and thermal planning, Edgecore attributes energy-efficiency gains to the 3 nm process and claims a “significant reduction in power consumption per gigabit,” adding that the platform “still can operate all 64 ports in 30W mode simultaneously.”

For software integration, Edgecore positions the switches for disaggregated deployments, stating support for “a broad ecosystem of open-source and commercial software.” It also lists precision timing support via Precision Time Protocol version 2 (PTPv2) and Synchronous Ethernet (SyncE), plus “deep telemetry,” to provide visibility into AI network operation.

“The transition to 1.6T Ethernet is a critical milestone in meeting the relentless bandwidth demands of next-generation AI clusters,” said Hasan Siraj, Vice President of Product Management, Core Switching Group, Broadcom. “By utilizing our latest Tomahawk 6 silicon, Edgecore is delivering a high-performance, power-efficient platform to the open networking ecosystem that empowers customers to scale their infrastructure with confidence. We are pleased to collaborate with Edgecore as they bring this advanced silicon intelligence to the open networking ecosystem, driving new levels of efficiency and performance in the modern data center.”

Source: Edgecore Networks

Chroma — below posts

Get Data Center Engineering News In Your Inbox:

By subscribing, you agree to our Privacy Policy and Terms of Use. You can unsubscribe at any time.

Sponsored: Infineon at OCP 2026: Exclusive Content Hub. Executive interviews, technical documents, reference designs, and live expert panel. Unlock the hub.
TDK — tower ()

Popular Posts:

Frontale_(sh)
Mitsubishi Electric launches 1.75 MW magnetic-bearing air-cooled chiller for AI data centers
3
3.2T near-packaged optics uses glass substrate to cut AI switch power 50%
tranetechnologies
Trane demos 800 VDC chiller for AI data centers, 3.5 MW cooling capacity
LiquidJet-Diamond-Oct-2026
LiquidJet Diamond coldplate claims 35% more AI tokens per watt
Giuseppe Bernacchia, distinguished engineer at Infineon
Infineon on protecting the 800 V rack: fast enough to save it, slow enough not to trip it

Share Your Data Center Engineering News

Do you have a new product announcement, webinar, whitepaper, or article topic? 

Chroma — tower

Get Data Center Engineering News In Your Inbox:

By subscribing, you agree to our Privacy Policy and Terms of Use. You can unsubscribe at any time.

TDK — leaderboard ()