What Is the Difference Between QSFP+, QSFP28, QSFP56, QSFP-DD?

QSFP+, QSFP28, QSFP56, and QSFP-DD are high-density pluggable transceiver form factors used in switches, routers, servers, and data center networks. Although they share the QSFP design family, they differ in lane count, signaling rate, aggregate bandwidth, and host-port compatibility. Understanding these differences helps network engineers select the right optical transceiver or direct attach cable (DAC) for a specific link.

In typical Ethernet deployments, QSFP+ is associated with 40G, QSFP28 with 100G, QSFP56 with 200G, and QSFP-DD with 400G. Newer QSFP-DD generations also support 800G and beyond. However, the form factor alone does not determine the complete optical specification, transmission distance, or supported application.

1. QSFP+ vs QSFP28 vs QSFP56 vs QSFP-DD: Key Differences

The lane configurations above describe common Ethernet implementations rather than every possible module. Optical lane arrangements, modulation, and breakout support depend on the transceiver specification and the equipment at both ends of the link.

2. What Is a QSFP+ Transceiver?

QSFP+ stands for Quad Small Form-factor Pluggable Plus. It uses four electrical lanes, commonly operating at approximately 10 Gbps per lane, to deliver an aggregate data rate of 40 Gbps.

2.1 Main Characteristics of QSFP+

• Typical bandwidth: 40 Gbps.

• Lane configuration: 4 × 10 Gbps.

• Common signaling: NRZ.

• Typical applications: 40G Ethernet switches, server connections, and data center interconnects designed for 40G links.

QSFP+ optical modules are available in different optical variants, including short-reach multimode and longer-reach single-mode options. The actual reach depends on the module, wavelength, fiber type, and link budget. QSFP+ DAC and AOC cables are also used for suitable short-distance connections.

2.2 When Should You Use QSFP+?

QSFP+ is suitable when the network equipment provides 40G ports and the required bandwidth does not justify upgrading to a higher-speed interface. It can also remain useful in established networks where replacing switches and cabling would be more expensive than maintaining the existing 40G infrastructure.

3. What Is a QSFP28 Transceiver?

QSFP28 retains the four-lane QSFP architecture but supports approximately 25 Gbps per electrical lane, enabling a typical aggregate rate of 100 Gbps. It became a widely used form factor for 100G Ethernet switches, routers, and data center links.

3.1 Main Characteristics of QSFP28

• Typical bandwidth: 100 Gbps.

• Lane configuration: 4 × 25 Gbps.

• Common signaling: NRZ for standard 100G 4 × 25G Ethernet implementations.

• Typical applications: 100G switch uplinks, leaf-spine networks, server connectivity, and data center aggregation.

Common QSFP28 optical options include SR4, PSM4, CWDM4, LR4, and DR-family modules. They differ in optical architecture, fiber requirements, connector type, and reach. The correct choice depends on the link distance and the installed fiber infrastructure.

3.2 QSFP28 Breakout Applications

A supported 100G QSFP28 port can sometimes be configured as four 25G links using an appropriate breakout cable or optical solution. This is useful when connecting a high-speed switch port to multiple 25G server or switch ports. The switch's port configuration, transceiver type, and cabling must all support the intended breakout arrangement.

4. What Is a QSFP56 Transceiver?

QSFP56 increases the data rate per lane while retaining four electrical lanes. In a common 200G Ethernet implementation, each lane operates at approximately 50 Gbps, delivering an aggregate bandwidth of 200 Gbps.

4.1 Main Characteristics of QSFP56

• Typical bandwidth: 200 Gbps.

• Lane configuration: 4 × 50 Gbps.

• Common signaling: PAM4 for 200G Ethernet implementations using 50G-per-lane signaling.

• Typical applications: 200G Ethernet switching, data center fabrics, and high-performance computing networks.

PAM4 encodes two bits per symbol, increasing the amount of data transmitted per symbol compared with NRZ. This enables higher lane rates without simply doubling the symbol rate, but it also introduces tighter signal-integrity and receiver-performance requirements.

4.2 QSFP56 vs QSFP28

The primary difference is the data rate per lane and the resulting aggregate bandwidth. A typical QSFP28 100G interface uses four 25G lanes, while a QSFP56 200G interface uses four 50G lanes. Although their physical form factors are closely related, their electrical signaling requirements and supported data rates differ.

A QSFP56 port may support selected lower-speed QSFP-family modules, but this is a host-device capability rather than a universal rule. Check the switch or network adapter's compatibility documentation before choosing a module.

5. What Is a QSFP-DD Transceiver?

QSFP-DD means Quad Small Form-factor Pluggable Double Density. It adds a second row of electrical contacts, increasing the number of high-speed electrical lanes from four to eight while retaining a compact QSFP-family form factor. This higher lane count allows substantially greater bandwidth per port.

5.1 Main Characteristics of QSFP-DD

• Common 400G configuration: 8 × 50 Gbps electrical lanes.

• Common 800G configuration: 8 × 100 Gbps electrical lanes in QSFP-DD800 implementations.

• Typical signaling: PAM4 for common 400G and 800G implementations.

• Typical applications: 400G and 800G switches, high-capacity data center fabrics, AI infrastructure, and cloud networking.

The original QSFP-DD generation supports up to 400 Gbps, while QSFP-DD800 increases the per-lane electrical rate to support up to 800 Gbps. Newer specifications extend the family further. When purchasing a module, identify the exact generation rather than assuming that every product labeled QSFP-DD supports the same speed.

5.2 QSFP-DD vs QSFP28

QSFP-DD provides eight high-speed electrical lanes instead of the four-lane architecture used by QSFP28. A common 100G QSFP28 module uses 4 × 25G lanes, whereas a 400G QSFP-DD module commonly uses 8 × 50G lanes. The additional contacts and higher signaling rates allow QSFP-DD systems to deliver more bandwidth per port.

QSFP-DD cages designed to the relevant specifications can support certain older QSFP modules, including QSFP+ and QSFP28. However, backward physical compatibility does not guarantee that every legacy module will operate at every rate or be supported by every switch. Consult the host platform's compatibility matrix.

6. QSFP+ vs QSFP28 vs QSFP56 vs QSFP-DD: Which Should You Choose?

Before selecting a transceiver, verify these factors:

1. Switch or NIC support: Confirm the port's supported form factor, data rates, and qualified transceiver list.

2. Optical reach: Select the module based on the required distance, wavelength, fiber type, and link budget.

3. Fiber and connector: Check whether the link uses duplex LC, MPO/MTP, or another specified interface.

4. Breakout requirements: Confirm the host supports the intended lane mapping and split configuration.

5. Thermal and power limits: Ensure the module fits the port's power budget and cooling design.

6. End-to-end compatibility: Check the transceivers, cables, port speeds, and configurations at both ends.

7. Frequently Asked Questions

8. Conclusion

QSFP+, QSFP28, QSFP56, and QSFP-DD represent different stages in the development of high-density network interfaces. QSFP+ commonly serves 40G links, QSFP28 supports 100G, QSFP56 supports 200G, and QSFP-DD enables 400G and higher-speed connections in newer generations. The most suitable option depends on the required bandwidth, host compatibility, optical reach, and network architecture—not simply the module's physical size.

References

• NVIDIA — Cable and Connector Definitions

• QSFP-DD MSA — Hardware Specification

• Cisco — QSFP-DD Optical Transceivers for High-Speed Connections

• SNIA — SFF QSFP+ Transceiver Solutions

• Cisco — What Are Breakouts and How Do You Use Them?