Driven by the increasing demand for bandwidth requirement, the 25G Gigabit Ethernet or 25GBASE-T standard has been released in recent years. Then 25G SFP28 assemblies like SFP28 transceiver and SFP28 DAC & AOC, come into being. Though the SFP28 assemblies have been used in a wide variety of high-density and low-power 25 Gigabit Ethernet connectivities, they are still different in optical communications. Why? The following part will uncover it.
To meet the growth for bandwidth in data centers, 25G SFP28 brings great benefits for system operators to achieve better network performance.
Before the 25GBASE-T standard arises, the Ethernet speed upgrade path was clearly defined as 10G – 40G -100G. However, the SFP28 provides a new way for server connections: 10G – 25G- 40G -100G. And since 25G offers a more efficient use of hardware and a more logical upgrade path to 100G, many companies like Cisco, Microsoft and Intel have put it into practice rapidly.
25G SFP28 takes existing module form factors, such as SFP28 and QSFP28, and allows for a breakout connection that is configurable as either 25G per lane or the full 100G without changing the port on the front of switches. For example, SFP28 twianx cable, also called SFP28 DAC, is a type of 25G SFP28 assembly. It has a similar structure with other DACs like SFP+ DAC or QSFP+ DAC whose connectors are not real optical transceiver. Due to this reason, SFP28 twinax cable is cheaper than SFP28 AOCs and SFP28 optical transceiver, which offers a cost-effective solution for high density, high speed I/O data center applications.
Furthermore, QSFP28 to SFP28 breakout cable provides various cabling choices for switches and servers. In the following figure, there are two different links between the switch and servers: 25G and 50G server links. Both of them use the 100G QSFP28 to SFP28 breakout cables.
SFP28 optics offer various options for data center and high-performance computing networks applications. However, when using them, many users still have questions about their compatibility and would like to compare them with existing optical modules. That’s one reason why it gets so much attention. This part will explain the differences between SFP28 and SFP, SFP28 and QSFP28.
The answer for this question is “yes”. As we all know, SFP+ form factor is defined for 10G speeds. And it was later adopted for 25G but the electrical interface had to be upgraded to handle 25Gbps per lane. The pinouts of SFP28 and SFP+ connectors are mating compatible. Therefore, SFP28 is backwards compatible with SFP+ ports and will work sufficiently. In addition, compared with SFP+, SFP28 has increased bandwidth, superior impedance control and less cross talk in network connections. And SFP28 copper cable has significantly greater bandwidth and lower loss than SFP+ version.
Though there is a number “28” in their names, there are differences between the SFP28 and QSFP28 modules. SFP28 modules support 25Gbps per lane. While QSFP28 supports four independent lanes, and each is 25Gbps. Both of them can be used in 100G networks, but the SFP28 is applied in the form of QSFP28 to SFP28 breakout cables.
As a new-comer, the compatibility and the benefits makes 25G SFP28 different from 10G and 40G in some degree. The advent of 25G changes the connection speed of servers, and the migration path will from 25G to 50G or 100G, which in turn accelerate the development of 25G. 25G SFP28 assemblies, as an important bridge in this progress, are booming now and in the near future.