800G/1.6T ACC Cables

Ultra-High Bandwidth for AI & HPC Fabrics

Overview

The NSComm 800G/1.6T Active Direct Attach Copper (ACC) collection is designed for next-generation AI, HPC, and hyperscale data centers that demand extreme bandwidth with minimal latency. These cables use active twinax copper with integrated QSFP-DD or OSFP connectors, delivering plug-and-play, short-reach interconnects without the complexity of optical modules and fiber patching.

This category covers 800G Ethernet and InfiniBand NDR/XDR scenarios, supporting common deployment models such as GPU server uplinks to ToR (leaf), leaf-to-spine, and switch-to-NIC connections inside high-density racks or across adjacent racks.

Broad Compatibility (NVIDIA-First at 800G/1.6T)

At ultra-high speeds, NVIDIA platforms dominate many AI fabrics. This collection focuses on NVIDIA/Mellanox-ready OSFP solutions, while also providing a generic QSFP-DD 800G ACC option for broader environments.

If you need a specific coding/profile for your switch or NIC, our engineers can help you match the right option quickly.

Product Range

A complete lineup of straight-through and breakout ACC cables for 800G/1.6T designs:

1. 800G QSFP-DD ACC (up to 2m)

A compact, generic 800G copper option for high-density in-rack links, where short reach and clean cabling are priorities.

2. 800G OSFP Finned Top ACC (up to 5m) - NVIDIA/Mellanox Focus

Built for high-power switch cages that benefit from finned-top cooling, commonly used for AI/HPC fabrics and cross-rack short runs.

3. 800G OSFP → 4×200G QSFP112 Breakout ACC (up to 5m) - NVIDIA/Mellanox Focus

Used to connect an 800G switch port to four 200G endpoints, ideal for AI pods where 200G NICs remain common.

4. 800G OSFP → 2×400G QSFP112 Breakout ACC (up to 5m) - NVIDIA/Mellanox Focus

A practical path to link an 800G switch to two 400G NICs/DPUs, often seen in upgraded GPU racks.

5. 800G OSFP → 2×400G OSFP Flat Top Breakout ACC (up to 5m) - NVIDIA/Mellanox Focus

Designed for connecting 800G OSFP switch ports to two 400G OSFP endpoints, supporting clean migration and high port efficiency in modern fabrics.

6. 1.6T OSFP Flat Top ACC (up to 3m) - NVIDIA/Mellanox Focus

For cutting-edge InfiniBand XDR and ultra-high bandwidth environments, supporting dense supercomputing and hyperscale designs where every hop matters.

Key Benefits

• Ultra-low latency: Direct electrical interconnect for AI training traffic.

• Simplified 800G/1.6T deployment: No separate optics or fiber patching.

• Breakout flexibility: Connect 800G ports to 400G/200G endpoints efficiently.

• AI/HPC ready: NVIDIA/Mellanox-oriented OSFP ecosystem support.

• Clean cabling for dense racks: Optimized for ToR uplinks and short interconnects.

Buying Guide

• Pick the connector family:

  • QSFP-DD for compact 800G in-rack use cases

  • OSFP for many NVIDIA-centric 800G/1.6T fabrics and high-power switch cages

• Choose straight vs breakout:

  • Straight 800G/1.6T for direct switch-to-switch or switch-to-NIC links

  • Breakout when your endpoints are 400G or 200G

• Match the scenario:

  • GPU servers → ToR (leaf): clean, short copper uplinks

  • Leaf ↔ spine / adjacent racks: high bandwidth with simple installation

  • Switch → multiple NICs: breakout for port density and phased upgrades

Conclusion

The 800G/1.6T Active Direct Attach Copper Cable Collection is built for the most demanding AI and HPC fabrics-combining extreme bandwidth, low latency, and breakout flexibility for real-world upgrade paths. Whether you're deploying 800G Ethernet today or moving toward 1.6T InfiniBand XDR, NSComm ACC helps you connect faster with less complexity.

⚡ Power your AI fabric with NSComm 800G/1.6T ACC-dense, low-latency, and upgrade-ready.