Extreme Networks 100GbE QSFP28 SR4 Optical Transceiver

The Extreme Networks 100GbE QSFP28 SR4 Optical Transceiver (part number 10502) is a high-density, short-reach optical module designed for 100 Gigabit Ethernet connectivity within modern data center environments. Built to the QSFP28 form factor, it aggregates four 25Gbps lanes over multimode fiber using parallel optics, conforming to the IEEE 802.3bm 100GBASE-SR4 standard. The module interfaces via an MPO-12 connector and operates over OM3 or OM4 multimode fiber, supporting link distances of up to 70 meters on OM3 and up to 100 meters on OM4 — making it well suited for intra-rack and inter-rack connections across top-of-rack switching and spine-leaf architectures.

Engineered for compatibility with Extreme Networks switching and routing platforms, the 10502 transceiver supports hot-swappable insertion without disrupting active network traffic, enabling non-disruptive capacity additions in high-availability environments. The module operates on a 3.3V supply and is designed to meet or conform to the QSFP28 MSA (Multi-Source Agreement) specification, ensuring consistent optical and electrical interface behavior. Low power consumption relative to longer-reach alternatives makes it a practical choice for high port-density deployments where thermal budget and power draw per rack unit are critical operational considerations.

As AI/ML training clusters and high-performance data center fabrics increasingly demand 100G inter-node bandwidth at scale, the 10502 has seen elevated adoption in GPU cluster interconnects, distributed storage networks, and high-throughput east-west traffic architectures. Its parallel SR4 optical design provides the throughput density necessary for leaf-spine fabrics underpinning modern AI infrastructure buildouts, and its compatibility within the Extreme Networks ecosystem allows for integrated management and diagnostics through Extreme's network management platforms.

Ideal for

• Leaf-to-spine 100GbE uplink connections within data center fabric deployments running Extreme Networks switching infrastructure
• GPU cluster interconnects for AI and machine learning training workloads requiring high-bandwidth, low-latency east-west traffic between compute nodes
• Top-of-rack server connectivity for high-density virtualization and cloud-native application environments demanding 100G server uplinks
• Distributed all-flash and NVMe-over-Fabric storage network interconnects where throughput and low latency are critical to storage performance
• Data center interconnect at the campus or building level across short multimode fiber runs within the OM3/OM4 distance envelope
• Network capacity expansion in existing 100G infrastructure where hot-swap, non-disruptive module insertion is required to maintain uptime SLAs

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.