800G OSFP to 4×200G QSFP112 Active Optical Cable (AOC), Armored – 850nm OM3

• Supports 800G to 4×200G breakout connectivity
• OSFP112 to 4×QSFP112 configuration
• 106.25Gbps PAM4 data rate per lane
• 8×106.25Gbps PAM4 transmitter and receiver channels at the 800G end
• 2×106.25Gbps PAM4 transmitter and receiver channels at each QSFP112 end
• Integrated 8-channel 850nm VCSEL and PIN photodetector arrays at the OSFP112 end
• 850nm multimode optical transmission
• OM3 fiber with transmission distances up to 30m
• Armored cable construction for enhanced mechanical protection
• DSP-based signal processing
• DDM/DOM supported

• 800G to 4×200G Ethernet breakout
• 800G data center switch-to-200G switch connectivity
• AI and GPU cluster interconnects
• High-performance computing (HPC)
• High-density Ethernet switching
• Data center spine-and-leaf architectures
• 800G network aggregation and breakout
• Short-reach interconnects between high-density network equipment
• Applications requiring enhanced cable mechanical protection

OEM and ODM services are available for the 800G OSFP to 4×200G QSFP112 AOC to meet specific equipment and deployment requirements. Customization options can include cable length, armored cable construction, labeling, EEPROM configuration, cable identification, packaging, and other assembly specifications.

For specialized or high-density deployments, the cable assembly can be adapted to customer-defined mechanical protection and cable-management requirements. Customized identification and programming can also be provided to support specific host platforms and network management systems.

How does the 800G-to-4×200G breakout architecture work?

The OSFP112 end provides a single 800G interface, while the four QSFP112 ends provide four independent 200G interfaces. This allows one high-bandwidth 800G switch port to connect with four separate 200G network ports through a single integrated optical cable assembly.

What is the role of the armored construction?

The armored design provides additional mechanical protection around the optical cable, helping improve resistance to external compression, bending, and handling during installation. It is particularly useful in dense networking environments where cables may be exposed to more demanding routing and management conditions.

Why does each QSFP112 end use two 106.25Gbps PAM4 lanes?

Each 200G QSFP112 end uses two high-speed PAM4 lanes, with each lane operating at 106.25Gbps. Four such 200G ends together provide the lane capacity required for the 800G OSFP112 side.

Why is the power consumption different between the OSFP112 and QSFP112 ends?

The OSFP112 end handles the full 800G interface and therefore contains more active optical and signal-processing functions. Each QSFP112 end handles a 200G connection, resulting in a lower power requirement. The specified maximum values are ≤14W for the OSFP112 end and ≤10W for each QSFP112 end.

What is the advantage of using an AOC for 800G breakout?

An AOC integrates the optical transmitters, receivers, and fiber assembly into one factory-terminated solution. This simplifies deployment compared with installing multiple optical transceivers and separate fiber patch cables, while also providing greater reach and routing flexibility than short-reach copper breakout cables.

Is this cable suitable for AI and GPU clusters?

Yes. The combination of 800G bandwidth, four 200G breakout links, low-latency optical transmission, and high-density OSFP112/QSFP112 connectivity makes it well suited to AI and GPU cluster fabrics, HPC environments, and other bandwidth-intensive data center architectures.