A 400G-to-4x100G breakout cable connects one 400G network port to four independent 100G ports. It allows a high-speed switch interface to communicate with multiple lower-speed devices through a single cable assembly, making it useful in data centers, AI infrastructure, high-performance computing (HPC), and enterprise networks.
For example, a 400G QSFP-DD port on a switch can connect to four 100G network interfaces using a compatible breakout cable. Each branch provides a separate 100G link, while the total aggregate bandwidth remains 400Gbps.
The operation involves more than physically dividing a cable into four branches. The host and branch interfaces must have compatible electrical or optical lane architectures. Depending on the design, the cable may use direct lane mapping or incorporate a Gearbox to convert between different signaling formats.
This article explains the internal architecture of 400G-to-4x100G breakout connections, how PAM4 and NRZ signaling interact, the differences between DAC and AOC solutions, and what to check before deployment.
1. What Is a 400G to 4x100G Breakout Cable?
A 400G-to-4x100G breakout cable is a high-speed interconnect assembly with one 400G connector at the host end and four 100G connectors at the branch end.
It distributes the bandwidth of one 400G interface across four separate 100G connections:
Host side: 1 × 400G port.
Branch side: 4 × 100G ports.
Aggregate bandwidth: 400Gbps.
Bandwidth per branch: 100Gbps.
A typical configuration uses a QSFP-DD connector at the 400G end and four compatible 100G connectors at the other end. The exact connector combination depends on the host platform, branch interfaces, and cable design.
1.1 Why Use a 400G-to-4x100G Connection?
Modern data center switches increasingly provide high-density 400G ports, while many servers, storage systems, and downstream switches still use 100G interfaces.
Instead of using four separate 100G host ports, a compatible 400G port can serve four 100G endpoints. This helps improve port utilization and provides a practical way to connect different generations of network equipment.
1.2 How Is Bandwidth Distributed?
The bandwidth is divided equally among the four branches:
4 × 100Gbps = 400Gbps
Each branch operates as an independent 100G connection. The cable does not create additional bandwidth; it distributes the available capacity of the original 400G interface.
2. How Does a 400G-to-4x100G Breakout Cable Work?
The basic operation involves three stages: receiving signals from the 400G host port, mapping or converting the signals into four 100G channels, and transmitting those channels through the four branch connectors.
2.1 Step 1: Receive the 400G Host Signal
In a typical 400G QSFP-DD implementation, the host electrical interface uses eight lanes operating at approximately 50Gbps per lane with PAM4 signaling.
The aggregate bandwidth is:
8 × 50Gbps = 400Gbps
These electrical lanes enter the breakout cable or its integrated active electronics, depending on the cable architecture.
2.2 Step 2: Map or Convert the Electrical Lanes
The eight high-speed lanes must be distributed into four groups, each supporting a 100G connection. How this happens depends on the signaling architecture of the branch ports.
There are two common scenarios:
Direct lane mapping: The host and branch interfaces use compatible electrical lane rates and signaling formats, allowing the lanes to be routed to the corresponding branch connectors.
Gearbox conversion: The host and branch interfaces use different lane rates or modulation formats. A Gearbox converts the incoming signal arrangement into the format required by the branch interfaces.
This distinction is particularly important when connecting a modern 400G PAM4 interface to traditional 100G QSFP28 ports.
2.3 Step 3: Distribute Signals to Four 100G Ports
After the required lane mapping or conversion, the signals are distributed across four branch connections. Each branch carries 100Gbps of aggregate bandwidth.
The four branch connectors can then connect to separate servers, network adapters, storage devices, or downstream switches.
2.4 Step 4: Establish Four Independent Network Links
The connected switch must support the 4x100G breakout mode. Its port configuration and software determine how the physical host interface is presented as four logical or channelized links.
Once the links are correctly configured and compatible devices are connected, each branch can operate independently. Traffic distribution across those links is handled by the network equipment and its forwarding configuration, not by the cable itself.
3. Understanding the Lane Architecture: 400G vs. 100G
To understand why some 400G-to-4x100G cables require a Gearbox, it is necessary to examine the electrical lane architecture at both ends.
3.1 400G QSFP-DD Electrical Lanes
A common 400G QSFP-DD implementation uses eight electrical lanes with PAM4 signaling. Each lane operates at approximately 50Gbps, producing 400Gbps of aggregate bandwidth.
PAM4 represents four signal levels, allowing two bits to be encoded per symbol. This makes it possible to transmit more data per symbol than traditional NRZ signaling.
3.2 Traditional 100G QSFP28 Electrical Lanes
A traditional 100G QSFP28 interface commonly uses four electrical lanes operating at approximately 25Gbps each with NRZ signaling.
Its aggregate bandwidth is:
4 × 25Gbps = 100Gbps
Four such 100G interfaces therefore require sixteen 25G NRZ electrical lanes in total.
3.3 Why Direct Lane Mapping May Not Be Enough
The 400G host side and four traditional QSFP28 branch interfaces use different electrical signaling arrangements:
400G host: 8 × approximately 50Gbps PAM4.
Four 100G QSFP28 branches: 16 × approximately 25Gbps NRZ.
Simply connecting eight PAM4 lanes to sixteen NRZ lanes is not a passive wiring operation. The signal format and lane arrangement must be converted.
A suitable Gearbox or active conversion architecture can perform this function, allowing the 400G host interface to communicate with four traditional 100G QSFP28 interfaces.
3.4 Not Every 100G Interface Uses the Same Lane Architecture
The 100G label describes the aggregate data rate, not necessarily the electrical lane arrangement. Some 100G interfaces use four 25G NRZ lanes, while other implementations use fewer higher-speed PAM4 lanes.
For example, selected 100G QSFP56-based InfiniBand interfaces use a different lane arrangement from traditional 100G QSFP28 Ethernet interfaces. Therefore, the correct breakout design must be selected according to the actual interface specification rather than the connector shape or data rate alone.
4. What Is the Role of a Gearbox in 400G-to-4x100G Breakout?
A Gearbox is an electronic circuit that converts between different lane counts or signaling rates. In selected 400G breakout cable designs, it enables compatibility between the 400G host interface and four 100G branch interfaces.
4.1 PAM4-to-NRZ Conversion
When a 400G QSFP-DD host using eight PAM4 electrical lanes connects to four traditional QSFP28 100G ports using sixteen NRZ lanes, a Gearbox may be used to convert the signaling arrangement.
In simplified terms, it performs the following functions:
Receives the high-speed electrical lanes from the 400G host.
Processes the incoming signals.
Converts the lane arrangement and signaling format as required.
Outputs the corresponding electrical signals for the four 100G branches.
4.2 Gearbox vs. Retimer
A Gearbox and a Retimer perform different functions, although both may appear in active cable designs.
| Feature | Gearbox | Retimer |
|---|---|---|
| Main Function | Converts lane rates or lane arrangements | Recovers and retransmits signals |
| Typical Role | Interface rate or format adaptation | Signal integrity improvement |
| Clock Recovery | Depends on implementation | Typically uses CDR |
| Lane Conversion | Yes, when designed for it | Not necessarily |
| Application | Different host and branch signaling architectures | Electrical channels requiring signal regeneration |
A Retimer does not automatically perform the same conversion as a Gearbox. Some active cable designs integrate both functions, while others use only one type of signal-processing circuit.
4.3 Does Every 400G-to-4x100G Cable Need a Gearbox?
No. A Gearbox is required only when the selected host and branch interfaces need lane-rate or signaling-format conversion.
If both ends use compatible electrical signaling and lane arrangements, direct lane mapping may be sufficient. The exact cable architecture should be confirmed through the manufacturer's technical specifications.
5. Different Types of 400G-to-4x100G Breakout Solutions
There are several ways to implement a 400G-to-4x100G connection. The main options include DAC, AEC, AOC, and optical module breakout.
5.1 400G-to-4x100G DAC Breakout Cable
A DAC (Direct Attach Copper) breakout cable uses copper twinaxial conductors to connect one 400G host port to four 100G branch ports.
Passive DAC is suitable for very short connections when the electrical lane architecture is compatible. If the host and branch interfaces require signaling conversion, an active cable design with appropriate electronics may be necessary.
Key characteristics:
Copper twinaxial transmission.
Very low power consumption for passive models.
Low latency.
Suitable for short rack-level connections.
Reach depends on cable design, wire gauge, and signaling requirements.
5.2 400G-to-4x100G AEC Breakout Cable
An AEC (Active Electrical Cable) uses copper conductors with integrated active electronics. Depending on the design, these electronics may include Retimers, Gearboxes, or other signal-conditioning functions.
AEC can support electrical connections that exceed the practical reach of some passive copper designs. It may also accommodate specific host-to-branch signaling requirements.
However, the presence of a Retimer does not automatically guarantee PAM4-to-NRZ conversion. The cable's actual conversion capability must be verified.
5.3 400G-to-4x100G AOC Breakout Cable
An AOC (Active Optical Cable) uses optical fiber and integrated optical components to connect one 400G interface to four 100G interfaces.
The host-side optical electronics convert electrical signals into optical signals. The optical signals travel through the fiber channels, and the branch-side electronics convert them back into electrical signals.
Depending on the interface architecture, the integrated electronics may also provide the required lane mapping or signal conversion.
Key characteristics:
Optical fiber transmission.
Integrated optical transmitters and receivers.
Longer reach than typical passive DAC.
Immunity to electromagnetic interference along the fiber.
Suitable for rack-to-rack and other short-reach optical connections.
5.4 400G Optical Module with Fiber Breakout
Another approach uses a 400G optical transceiver at the host end, four separate 100G optical transceivers at the branch ends, and a suitable fiber breakout harness.
For example, a compatible 400GBASE-DR4 optical interface can connect to four 100GBASE-DR1 interfaces through a suitable MPO-to-four-duplex-LC fiber arrangement.
This is different from an integrated AOC. The optical modules and fiber cable are separate components, allowing individual modules or fiber assemblies to be replaced independently.
6. 400G-to-4x100G DAC vs. AOC vs. AEC
The choice between DAC, AEC, and AOC depends on the required reach, power consumption, cable routing, and electrical interface architecture.
| Feature | DAC | AEC | AOC |
|---|---|---|---|
| Transmission Medium | Copper twinax | Copper twinax | Optical fiber |
| Active Electronics | Not in passive models | Yes | Yes |
| Signal Conditioning | Passive path or active design | Retimer or other active circuitry, depending on model | Integrated optical conversion |
| Gearbox Support | Only in suitable active designs | Depends on model | Depends on optical engine design |
| Power Consumption | Lowest for passive models | Higher than passive DAC | Requires power for optical components |
| EMI Immunity | Lower | Lower | Optical fiber is immune to EMI |
| Typical Deployment | Very short links | Short electrical links requiring active processing | Longer short-reach optical links |
There is no universal reach limit for all three cable types. Supported distances depend on the specific product, signaling rate, cable construction, and equipment compatibility.
7. 400G-to-4x100G Breakout Cable Applications
7.1 Spine-to-Leaf Switch Connections
In a leaf-spine architecture, spine switches may provide 400G ports while leaf switches use 100G uplinks. A 400G-to-4x100G breakout cable allows one spine port to connect to four compatible 100G leaf ports.
This provides a practical way to deploy higher-speed spine infrastructure while retaining existing 100G switching equipment.
7.2 Switch-to-Server Connectivity
In high-density data centers, one 400G switch port can connect to four separate 100G server interfaces. This is useful when multiple servers require individual 100G connections to the same switching layer.
7.3 AI and GPU Cluster Networks
AI clusters rely on high-bandwidth communication between GPU servers and network switches. A 400G-to-4x100G breakout configuration can connect high-speed switching equipment to multiple compatible 100G network interfaces.
The suitability of this configuration depends on the actual network topology and bandwidth requirements of the computing cluster.
7.4 Storage Network Connectivity
Distributed storage systems and NVMe-based infrastructure require high-throughput connections between switches and storage nodes. Breakout cables can provide multiple 100G links from one 400G port when the connected devices support the required interfaces.
7.5 100G-to-400G Network Migration
During a network upgrade, operators may install 400G switches before replacing all existing 100G devices. A compatible breakout cable allows the new 400G switching layer to connect to four 100G endpoints, supporting a gradual infrastructure migration.
8. How to Configure a 400G Port for 4x100G Breakout
Installing the cable is only part of the process. The switch must also support and be configured for the required breakout mode.
8.1 Verify Switch Port Compatibility
Check whether the switch supports 400G-to-4x100G channelization on the selected port. Not every 400G port or transceiver supports every breakout configuration.
8.2 Check Cable and Interface Compatibility
Verify the following information:
400G host connector and form factor.
Four branch connector types.
Host and branch electrical lane architecture.
PAM4 or NRZ signaling requirements.
Required Gearbox or active signal-conditioning functions.
Supported Ethernet or InfiniBand protocol.
8.3 Configure the Breakout Mode
Use the switch operating system or management interface to configure the host port for four 100G connections. The command syntax varies by vendor and switch model.
Some platforms require a port-profile or interface-channelization setting, while others provide breakout configuration through a dedicated interface command.
8.4 Connect the Four Branch Ports
Insert the 400G connector into the host port and connect each branch connector to its corresponding 100G device or switch port.
For optical connections, follow the manufacturer's handling and cleaning recommendations for optical interfaces.
8.5 Verify Link Status and Configuration
After installation, confirm that all four branch links are recognized and operating at the expected speed. Check the switch interface status, transceiver information, error counters, and supported link configuration.
If one or more links fail to establish, verify the breakout configuration, cable compatibility, lane mapping, and endpoint settings.
9. Common Problems with 400G-to-4x100G Breakout Connections
9.1 The Switch Does Not Support 4x100G Breakout
A physical breakout cable cannot create a breakout mode that the switch does not support. Confirm port channelization capabilities in the equipment documentation.
9.2 PAM4 and NRZ Interfaces Are Incompatible
A 400G PAM4 host interface and four traditional 100G NRZ interfaces may require a suitable Gearbox. A passive cable that only rearranges electrical lanes cannot automatically perform this conversion.
9.3 Incorrect Branch Connector Selection
QSFP28, QSFP56, QSFP-DD, and other connector types have different mechanical and electrical characteristics. Select the branch connector according to the actual 100G device interface.
9.4 Cable Protocol Mismatch
Ethernet and InfiniBand products may use different electrical specifications, coding, and compatibility requirements. Confirm that the selected cable is designed for the intended network protocol.
9.5 Incompatible Cable Length
A cable may support the correct data rate and connector arrangement but still fail to meet the required distance. Check the manufacturer's supported reach for the exact cable model.
9.6 Incorrect Switch Port Configuration
If the host port remains in native 400G mode instead of the required breakout mode, the four branch connections may not establish correctly. Confirm that the switch configuration matches the physical cable architecture.
10. C-LIGHT 400G-to-4x100G Breakout Cable Solutions
C-LIGHT provides 400G DAC configurations for high-speed data center and InfiniBand environments, including selected 400G-to-4x100G breakout designs.
One listed configuration is the 400G QSFP-DD-to-4x100G QSFP56 DAC, designed for compatible 400G-to-100G InfiniBand connections.
| Parameter | C-LIGHT Configuration |
|---|---|
| Host Interface | 400G QSFP-DD |
| Branch Interfaces | 4 × 100G QSFP56 |
| Breakout Ratio | 1 × 400G to 4 × 100G |
| Part Number | CLQD4-4Q1CU-0.5~3 |
| Transmission Medium | Copper twinax |
| Listed Cable Length | 0.5m–3m |
| Wire Gauge | 30/28/26 AWG options |
| Listed Application | 400G / 100G InfiniBand |
This product is specifically listed as a 400G-to-4x100G InfiniBand configuration. It should not be assumed to support every 400G Ethernet-to-100G QSFP28 combination. Confirm the host and branch interface specifications before selecting it.
For additional configurations, visit the C-LIGHT 400G DAC product page.
For optical connectivity requirements, see the C-LIGHT 400G AOC product family. The exact breakout connector combination and protocol support should be confirmed against the selected model.
11. Frequently Asked Questions About 400G-to-4x100G Breakout Cables
Q1: What does 400G to 4x100G mean?
It means one 400G host port is connected to four independent 100G ports, providing 400Gbps of aggregate bandwidth.
Q2: How does a 400G-to-4x100G breakout cable work?
It maps or converts the high-speed host signals into four 100G branch connections through compatible electrical or optical channels.
Q3: Does a 400G-to-4x100G cable need a Gearbox?
Not always. A Gearbox is needed when the host and branch interfaces require different lane rates or signaling formats, such as certain PAM4-to-NRZ configurations.
Q4: What is the difference between PAM4 and NRZ in 400G breakout?
PAM4 carries two bits per symbol using four signal levels, while NRZ uses two signal levels to carry one bit per symbol. The two formats require compatible interface or conversion circuitry.
Q5: Can a 400G QSFP-DD port connect to four 100G QSFP28 ports?
Yes, if the switch supports 4x100G breakout and the selected cable provides the required electrical signaling conversion and connector compatibility.
Q6: What is the difference between 400G DAC and 400G AOC breakout cables?
DAC uses copper twinax conductors, while AOC uses optical fiber and integrated optical components. DAC is generally suited to very short links, whereas AOC supports longer short-reach optical connections.
Q7: Does a 400G-to-4x100G breakout cable increase bandwidth?
No. It distributes the existing 400Gbps capacity across four 100Gbps links without increasing the aggregate bandwidth.
Q8: Can a 400G optical module connect to four 100G optical modules?
Yes. Compatible 400G and 100G optical interfaces can use a suitable fiber breakout harness, such as an MPO-to-four-duplex-LC arrangement for supported module combinations.
Q9: Does every 400G switch support 4x100G breakout?
No. Breakout support depends on the switch ASIC, port configuration, transceiver or cable compatibility, and software capabilities.
Q10: What should I check before buying a 400G-to-4x100G breakout cable?
Check the host connector, four branch connectors, signaling format, lane mapping, protocol, supported reach, and switch breakout capability.
12. Conclusion
A 400G-to-4x100G breakout cable allows one high-speed 400G interface to connect to four independent 100G ports. It is an effective interconnection option for data centers that need to connect 400G switching infrastructure with multiple 100G servers, network adapters, or downstream switches.
The key to understanding its operation is the lane architecture. A typical 400G QSFP-DD interface uses eight approximately 50Gbps PAM4 electrical lanes, while traditional 100G QSFP28 interfaces commonly use four approximately 25Gbps NRZ lanes per port. When these architectures differ, a suitable Gearbox or active conversion design may be required.
DAC, AEC, AOC, and optical module breakout solutions each address different deployment requirements. The right choice depends on transmission distance, signaling compatibility, cable management, and the supported breakout modes of the network equipment.
For reliable deployment, always verify the electrical or optical lane mapping and protocol compatibility rather than selecting a cable based only on its 400G and 100G labels.
Explore C-LIGHT 400G DAC and 400G AOC solutions for high-speed data center interconnection requirements.
TEL:+86 132 6656 7067




















































>
>
>
>
>
>
>
>