A Comprehensive Guide to 400G Ethernet DAC Cables

With the introduction and gradual adoption of 400G networks, the demand in this area is continuously increasing. Gradually, high-bandwidth demand scenarios are no longer limited to large-scale data center deployments or inter-building connections. Short-distance links may also require 400G deployment, and at this time, 400G Ethernet DAC cables are emerging. As a mature low-power integrated solution recognized by the market, DAC maintains low-latency stability and has also been widely deployed in low-speed networks (such as 10G and 25G).

Meanwhile, 400G Ethernet DAC carries higher signal rates over limited copper media, and its underlying technology relies on PAM4 modulation. It increases the per-lane rate to 50Gbps or higher, and achieves a total bandwidth of 400Gbps through multi-lane parallel transmission. This type of cable, which integrates cable and module, is commonly used in various fields, including data centers, HPC, and AI GPU clusters. Configuring a 400G Ethernet DAC cable for your fixed switch becomes a convenient and reliable choice.

What is a 400G Ethernet DAC

The 400G Ethernet DAC is a short-range Ethernet technology based on high-speed copper interconnects for data transmission at 400 Gbps. This technology eliminates the need for electro-optical conversions, reducing power consumption for data transmission. This technology is used for inter-cabinet connections within the same rack or between two adjacent racks. The mainstream forms are currently OSFP, QSFP-DD, and QSFP112.

How 400G DAC Works

Like many high-speed modules, the 400G Ethernet DAC relies on a multi-lane parallel transmission architecture. It is typically composed of four or eight high-speed electrical lanes, each operating at approximately 50 Gbps. PAM4 modulation enables higher data-carrying capacity. After SerDes processing, the high-speed electrical signals enter the DAC cable and are transmitted in the form of differential signals over Twinax copper cables, achieving a total bandwidth of 400Gbps.

PAM4 signals are more prone to noise than traditional NRZ signals at high speeds. There is a greater risk of signal attenuation and crosstalk when transmitted over copper wires. This creates a greater need for signal integrity in 400G DAC, which brings forward the concept of FEC (Forward Error Correction). At the transmitting end, redundant information is added to the data. At the receiving end, algorithms automatically correct errors that occur during transmission. This reduces the bit error rate. Moreover, in Active DAC, there are circuits within the device for equalization, amplification, and pre-emphasis of the signals to counter high-frequency losses. It can be used as a means to increase stability.

Types of 400G DAC

Although all are 400G DACs, they are actually divided into multiple types based on structure and signal processing methods. In real deployment, it is necessary to select suitable 400G Ethernet DAC cables according to specific usage requirements.

Passive DAC

This type of DAC adopts a fully passive structure, and the structure is simple. So it does not include any signal amplification or processing circuits. However, in the 400G Ethernet environment, PAM4 signals require very high channel quality. So the simple structure limits the usable distance of the Passive DAC to only 0.5m to 2m. Its advantages lie in the lowest power consumption, latency, and cost. It is a good choice for scenarios that only require intra-rack connections.

Active DAC

Compared with a passive DAC, this type of DAC integrates active circuits inside the connectors and performs equalization and amplification on high-speed PAM4 signals. At the 400G transmission rate, this signal compensation mechanism effectively reduces the bit error rate and allows longer transmission distances. This enables it to support connections up to around 3m or even longer. For inter-rack connections, Active DAC usually performs better in terms of link stability.

Breakout DAC

This is a special type of DAC cable. It splits one 400G port into multiple lower-speed ports (such as 2×200G or 4×100G), and achieves multi-port output through internal wiring distribution. Breakout DAC effectively improves port utilization. You do not need to purchase new servers or devices when you need four 100G ports. One Breakout DAC Cable can achieve this requirement.

Key Features of 400G DAC

As one of the deployment options for 400G, 400G Ethernet DAC cables have some prominent features compared to other solutions (such as AOC and optical modules). Understanding these features allows you to make more confident decisions when selecting solutions.

• High bandwidth capability: Like other 400G high-speed modules, 400G Ethernet DAC achieves 400Gbps high-speed transmission capability based on multi-lane PAM4 technology. Its transmission capability can fully meet the requirements of AI computing, cloud data centers, and high-performance networks.
• Simple and reliable deployment: The integrated design of DAC cables simplifies the original combination of modules and patch cords, and it is designed as a unified module. It can achieve plug-and-play during use, and fully tested, reliable modules directly meet your needs.
• Low power consumption, low latency, and low cost: Compared with optical module solutions, DAC cables eliminate the optical-electrical conversion process, allowing devices to maintain lower power consumption at 400G rates. At the same time, the direct electrical transmission characteristic greatly reduces link latency. The simple structure and integrated design also reduce hardware costs, making it one of the most cost-effective solutions in 400G networks.
• Standardization and compatibility: 400G Ethernet DAC cables follow Ethernet standards and support interoperability between multi-vendor devices. Compared with InfiniBand DAC cables, Ethernet DAC cables have better openness and compatibility.

400G Ethernet DAC Application Scenarios

In actual deployment, 400G Ethernet DAC cables are mainly used for short-distance high-bandwidth interconnection scenarios within data centers. Connections between servers and switches,high-speed links between adjacent racks. In these scenarios, the link distance is usually within a few meters, where DAC can provide stable 400 Gbps bandwidth without optical-electrical conversion. Therefore, a 400G Ethernet DAC becomes a common choice in ToR (Top of Rack) architectures. In Ethernet-based AI clusters and RoCE networks, 400G DAC is also widely used for interconnects between GPUs or between GPUs and switches to meet high-throughput, low-latency data communication requirements.

At the same time, in the case of high-density switching networks, Breakout DAC enables the splitting of the 400G ports into multiple 100G or 200G ports. This helps in the utilization of the ports more efficiently, allowing the network architecture to support devices with different rates. As such, in the case where the distances are not significant, but the need for high-bandwidth network performance is significant, the 400G Ethernet DAC is an ideal option.

FAQs

#1 400G DAC vs AOC: which one should I choose?
For short distances, DAC is the most cost-effective choice. When meeting longer distances, AOC is better.

#2 Can you use only some branches of a breakout DAC?
Yes, the device will recognize each branch as an independent port, and unused branches will not affect other links. Four branches represent the maximum usage.

Conclusion

400G Ethernet DAC cables are an economical and efficient solution for achieving high bandwidth over short distances. Different types of DAC cables each have their own advantages. Understanding the advantages and limitations of these cables helps you better optimize data center cabling according to technical scenarios.

Read more

• DAC vs AOC vs Fiber: Understanding the Key Differences in Data Center Connectivity
• What is Breakout Cable? Everything You Need to Know
• Passive vs Active DAC cable: How to Choose?