Explore NADDOD Data Center Interconnect (DCI) Solutions

NADDOD's DCI interconnect solutions offer a comprehensive portfolio of 800G/400G/100G optical modules, meeting the interconnect needs of enterprise-level and hyperscale data centers. With high integration, reliability, and flexible scalability, it enables the construction of efficient, stable, and easily maintainable DCI network operations.

Data Center Interconnect (DCI) Overview

What is DCI?

Data Center Interconnect (DCI) refers to high-speed optical transmission links between data centers, whether within a campus, across metropolitan areas, or spanning regions. DCI enables distributed data centers to achieve data synchronization, business continuity, disaster recovery, and cross-domain collaboration for AI workloads and cloud services.

Why do we need DCI?

With the explosive growth of cloud computing, AI training, video streaming, and big data, a single data center can no longer meet all computing and storage demands. Enterprises and hyperscale operators rely on DCI to:

Ensure Business Continuity: Support cross-regional disaster recovery and high-availability architectures.
Improve Resource Utilization: Enable compute, storage, and network resource sharing across data centers.
Meet Escalating Bandwidth Demands: Deliver ultra-high-capacity transmission at the Tbps level.
In-depth optimization of hardware management, such as power supply and heat dissipation.

Why do we need DCI?

Ensure Business Continuity: Support cross-regional disaster recovery and high-availability architectures.
Improve Resource Utilization: Enable compute, storage, and network resource sharing across data centers.
Meet Escalating Bandwidth Demands: Deliver ultra-high-capacity transmission at the Tbps level.
In-depth optimization of hardware management, such as power supply and heat dissipation.

What are the challenges for DCI?

While DCI has become critical to modern data center architectures, it still faces several challenges:

Rapidly Growing Bandwidth Demand: AI and cloud applications are driving continuous growth in link capacity.
Transmission Distance and Performance: Maintaining low latency and high reliability over tens to hundreds of kilometers is required.
Power Consumption and Cost Pressure: High-speed interconnects impose stricter requirements on energy efficiency and cost per bit.
Rapid Deployment and Scalability: The fast evolution of data center workloads demands networks with flexible and scalable capabilities.

What are the challenges for DCI?

While DCI has become critical to modern data center architectures, it still faces several challenges:

Transmission Distance and Performance: Maintaining low latency and high reliability over tens to hundreds of kilometers is required.
Power Consumption and Cost Pressure: High-speed interconnects impose stricter requirements on energy efficiency and cost per bit.
Rapid Deployment and Scalability: The fast evolution of data center workloads demands networks with flexible and scalable capabilities.

NADDOD Product Overview

Direct Detect Transceivers

Direct Detect Transceivers adopt PAM4 modulation and Intensity Modulation-Direct Detection (IM-DD) technology. Key advantages include simple architecture, low power consumption, and cost efficiency. The product series covers 100G/400G/800G with transmission distances from 2 km to 80 km, used in data center interconnects, campus,and short- to medium-reach DCI applications.

100G DCI Transceivers

(LR4/ER4/ZR4)

Built on a 4×25G CWDM design with QSFP28/LC duplex interfaces, it delivers reliable, low-power point-to-point transmission over 10–80 km, making it a mainstream solution for metro and long-reach data center interconnects.

400G DCI Transceivers

(FR4/LR4/ER4)

Designed to meet interconnection demands of AI clusters and large-scale data center Spine-Leaf architectures, balancing power consumption and transmission distance.

800G DCI Transceivers

(FR4/LR4)

Powered by a highly integrated 7nm DSP and low-power PAM4 driver, supporting high-speed interconnects over 2-10 km. Fully compatible with NVIDIA Spectrum-XGS platforms, positioned as a core solution for next-generation DCI deployments.

Digital Coherent Optics (DCO) Transceivers

Coherent optical modules employ PDM (polarization-division multiplexing), high-order QAM modulation formats, and advanced DSP to enable ultra-long-reach, high-speed data transmission. NADDOD's 400G ZR series (including ZR, ZR+, and ZRP+) supports transmission distances over 80 km and up to 450 km+ with optical amplification, combining ultra-long reach, high power efficiency, and optimized cost performance for high-capacity interconnect deployment.

NADDOD DCI Solutions

Transmission Distance	Transceivers	Applicable Scenarios	Selection Rationale	Customer Value Proposition
Within 2km	QDD-400G-FR4 OSFP-800G-2xFR4	Same campus data rooms, intra-building DCI, AI cluster interconnection	Low cost, low power, high maturity, meets high-density interconnect requirements	Reduces overall construction cost, simplifies deployment, ideal for high-port-density rack interconnect
2-10km	QSFP-100G-LR4 QDD-400G-LR4 QDD-400G-LR8 OSFP-800G-2xLR4 OSFP-800G-2xFR4L	Interconnection between data centers across buildings or campuses	IM-DD module supports up to 10km transmission, more cost-effective than coherent solutions	Delivers optimal TCO for medium-distance DCI interconnect
10-40km	QSFP-100G-ER4 QDD-400G-ER4 QDD-400G-ER8	DCI access layer, metro aggregation node interconnect	PAM4 solution retains cost and power efficiency advantages, supports 40km direct connection	Maintains a simple architecture while enabling longer reach, easier maintenance, and flexible scalability
40-80km	QSFP-100G-ZR4 QDD-400G-ZR QDD-400G-ZRP+	Direct data center interconnection, Inter-campus connectivity	Coherent modules fit 40-80km applications, balancing bandwidth and distance	Improves link stability, reduces repeaters, lowers network latency and power consumption
80-1000km	QSFP-100G-ZR4 QDD-400G-ZR QDD-400G-ZR+	Metro DCI, backbone access layer, Regional data center backbone interconnect	DCO coherent modules provide full optical-layer regeneration and ultra-long reach, alternative to traditional OTN platforms	Enables DCI-backbone convergence, high reliability, high bandwidth, and low O&M cost

Optical Transceivers Selection Solutions By Transmission Distance

Download NADDOD Solutions

Fully Tested and Compatible with Top-tier Coherent Switches

NADDOD's interconnection solutions undergo rigorous testing to ensure full compatibility with leading platforms such as Cisco NCS1000 series, Juniper routers and switches, Arista 7800R3 series, and Nokia 7250 IXR. This broad interoperability enables seamless deployment across diverse DCI environments, delivering high bandwidth, low latency, and reliable performance for large-scale data center interconnect applications.

Download Compatibility Matrix

NADDOD DCI Advantages

Direct Detect Transceivers

Business Value

Scalability Value

Technical Value

For AI and HPC clusters

Serves as the interconnection foundation for AI and HPC clusters, enabling high-speed data exchange and synchronization for accelerated model training and data flow.

Reduce Network Construction and O&M Costs

Modular, pluggable design eliminates the need for traditional OLS systems, significantly reducing rack space, equipment, and power costs.

High Interoperability and Open Standards

Compliant with IEEE / OIF 400ZR/ZR+ / QSFP-DD / OSFP standards to enable seamless multi-vendor interoperability and deployment.

For AI and HPC clusters

Serves as the interconnection foundation for AI and HPC clusters, enabling high-speed data exchange and synchronization for accelerated model training and data flow.

Reduce Network Construction and O&M Costs

Modular, pluggable design eliminates the need for traditional OLS systems, significantly reducing rack space, equipment, and power costs.

High Interoperability and Open Standards

Compliant with IEEE / OIF 400ZR/ZR+ / QSFP-DD / OSFP standards to enable seamless multi-vendor interoperability and deployment.

Digital Coherent Optics vs DWDM Transponders

Digital coherent optics solutions boost spectral efficiency on existing fiber, enabling rapid bandwidth expansion without new infrastructure. Pluggable coherent modules integrate directly into switch or router line cards, allowing space-free, cable-free upgrades. With higher energy efficiency, lower power and cooling demand, and reduced amplifier needs, coherent technology streamlines network design, cuts CAPEX/OPEX, and boosts deployment agility.

Digital Coherent Optics vs DWDM Transponders

DCI DCO Solutions vs Traditional DWDM

In traditional optical transport systems, DWDM modulation typically relies on standalone transponders, resulting in a more complex architecture and greater equipment footprint. In contrast, DCO (Digital Coherent Optics) modules used in DCI applications can directly generate coherent DWDM signals and interface with an OLS (Optical Line System) through single-mode fiber. This eliminates the need for additional modulation and forwarding equipment, simplifying the network and reducing overall cost.

DCI DCO Solutions vs Traditional DWDM

Next-Generation Network Silicon Accelerates DCI Evolution

As data center scale and AI/cloud application demands continue to grow, traditional network architectures face challenges in bandwidth, latency, and power consumption. The latest generation of network silicon—NVIDIA Spectrum-XGS, Cisco Silicon One P200, and Broadcom Jericho4—delivers higher throughput, flexible multi-rate support, low-latency processing, and power optimization, driving the evolution of cross-data center interconnect (DCI) technologies. NADDOD's 800G DCI solutions are tightly integrated with these high-performance chips, aiming to build DCI networks on efficient, low-power, secure, and future-ready infrastructure, helping ensure that you are ready for both today's and tomorrow's AI demands.

Extreme Bandwidth and Multi-Rate Flexibility

NVIDIA Spectrum-XGS and the Cisco Silicon One P200 support multi-rate interfaces ranging from 10G to 800G, while Broadcom Jericho4 offers high- density 400G interfaceswith scalable SerDes, enabling high - capacity transmission across data center interconnects and campus backbones.

Latency Optimization and Intelligent Traffic Management

NVIDIA Spectrum-XGS delivers fine-grained latency control through topology-aware congestion management and end-to-end telemetry. Cisco Silicon One P200 ensures deterministic performance under bursty traffic through deep buffering and enhanced clock synchronization, while Broadcom Jericho4 integrates deep buffers, RoCE-based lossless transport, and its Elastic Pipe architecture to provide highly efficient, low-jitter traffic scheduling.

Programmability and Cross-Domain Optimization

Modern silicon architectures are increasingly software-programmable and optimized for cross-data-center coordination. NVIDIA Spectrum-XGS enables end-to-end optimization across switches, smart NICs, and network operating systems, Cisco Silicon One P200 adopts a modular architecture with flexible SerDes configuration, and the Broadcom Jericho4 supports hybrid interfaces and programmable pipelines, making it suitable for WAN, Metro, DC deployments.

High Performance with Power Efficiency

NVIDIA Spectrum-XGS improves performance-per-watt across AI fabric deployments, the Cisco Silicon One P200 platform provides a power-optimized 51.2 Tbps full-duplex architecture, and Broadcom Jericho4 achieves up to 40% lower power per bit, delivering high performance with significantly reduced energy consumption.

Contact Us

Inquire about the NADDOD optical transceivers to customize your DCI solutions.

First Name*

Last Name*

Email Address*

Company/Organization

Country/Region*

Phone Number*

Your Inquiry*

I agree to NADDOD’s Privacy Policy and Terms of Use

Submit