Free Worldwide Shipping on Orders over $ 99
FAQ Warranty Payment Methods Shipping & Delivery Return Policy
Free Tech Support Demo Test Solution Design
Request a Sample Bulk Quotation Request Global Local Service
About Us NADDOD Overview NADDOD Certified
Contact Us sales@naddod.com +1 (888) 998 7997
InfiniBand Network

InfiniBand Network

 RoCE Network

RoCE Network

 Immersion Cooling Network

Immersion Cooling Network
RoCE Networking
Optical Transceivers
AOC, DAC, ACC & AEC
Switches & NICs
Fiber & Copper Cables

800G OSFP to 2x400G OSFP DAC Splitters

Filter

Product Category
Length
2 Results
Filter
Filter

Filter

Product Category
Length

Product Overview

NADDOD 800G breakout InfiniBand NDR DAC includes 800G Twin-port 2x400Gb/s OSFP to 2x400Gb/s OSFP InfiniBand NDR passive and active copper splitter cables. It provides connectivity between system units with an 800Gb/s OSFP port on one side and two 400Gb/s OSFP ports on the other. It is a high-quality, cost-effective alternative to fiber optics in 800Gb/s to 400Gb/s applications for high-speed, efficient, and sustainable network connectivity.

Related Blogs

NADDOD

InfiniBand Simplified: Core Technology FAQs

Learn the essential details about InfiniBand technology with this FAQ guide. Explore topics like hardware compatibility, best practices for connecting InfiniBand devices, and how to optimize your network for AI and HPC environments.
NADDOD

Spectrum-X: NVIDIA's Answer to AI Ethernet Challenges

In this article, we’ll explore the limitations of traditional Ethernet in handling AI traffic—focusing on issues like load imbalance, high latency, and poor congestion control, and also examine how Spectrum-X, with technologies like RoCE adaptive routing and NVIDIA Direct Data Placement, solves these challenges and enhances isolation performance in AI cloud environments.
NADDOD

Why AI/ML Networks Rely on RDMA?

RDMA plays a vital role in AI/ML deployments by enabling faster data transfers, reducing latency, and minimizing CPU load. Its integration into InfiniBand and Ethernet networks supports efficient AI/ML infrastructures, powering applications like training, inference, and reinforcement learning while enhancing overall network performance.
See more