As the computer industry grows, data centers are going through a constant evolution to meet the increasing data management demand.  Data centers have to change to accommodate the new designs, increased complexity, higher speeds, and higher densities.

Server racks were initially designed to support early data center design and development. However, the constant changes lead to the requirement of creating a common connector interface or standard. As there were various interfaces from multiple suppliers and hardware manufacturers, standardization was indeed a challenge.

SFF Committee and MSA

The Small Form Factor (SFF) committee took the initiative in creating the multi-source agreement (MSA) that created a standard guideline for different connector manufacturers, to design and produce components that can operate with each other.

Small form-factor pluggable (SFP) devices were the first interface, which was developed under the MSA agreement. Later the SFP standards were upgraded to SFP+. This helped to accommodate higher data transmission speed up to 10Gb/s per lane and to include the use of EEPROM.  These enhancements, however, did not contribute to overcome the limitations of single-lane interfaces.

QSFP – The 100Gb/s Milestone

Eventually with the data speeds increased up to 25Gb/s, demanding better bandwidth. This led to the implementation of the 4-lane Quad Small Form-factor Pluggable (QSFP) which supported 10Gb/s, or 40Gb/s aggregate bandwidth per lane. The technological revolutions at the time were rapid, and the demand for more advanced systems increased, which led to the development of 25Gb/s lane signal speeds with 100Gb/s aggregate bandwidth capability. The interfaces underwent several changes to maintain high-performance signal integrity, EMI shielding, and PCB footprint compatibility.

OSFP- Meeting New Challenges

After the success of QSFP, the next step was to aggregate bandwidth capability to 200Gb/s for which the Octal (8x) Small Form-factor Pluggable (OSFP) eight-lane interface was developed, with 25Gb/s of continued channel support per lane. The advanced OSFP connector design incorporates, 60 contacts/port on a 0.6mm pitch, with two rows of 16 high-speed differential pairs and 10 power/control contacts. This system will also support the PAM-4 (pulse amplitude modulation) communications protocol that helps to double the connection bandwidth of the eight, 25Gb/s ports to 50Gb/s, facilitating a total data transmission rate of 400Gb/s.

QSFP DD– Evolving Advancements

400Gb/s is the next milestone requiring, a new method of doubling the density of the standard 100Gb/s QSFP module. The Double Density or QSFP DD could accommodate 8 lanes of 25Gb/s data transmission (200Gb/s) or, 8 lanes of 50Gb/s (400Gb/s) using PAM-4 modulation in a single QSFP DD port. This interface can double the bandwidth within the given space and are backward pluggable with existing QSFP hardware.

RCx advances intra-rack connections

The evolution of data communication standards has now moved even further with the introduction of RCx interfaces that are specifically designed to provide 3-meter short-run intra-rack connectivity for 25Gb/s, 50Gb/s, and 100Gb/s Ethernet applications. These offer a more cost-effective, simple and flexible interconnect system. Compared to legacy intra-rack solutions, like 100G QSFP and SFP28 interfaces, these use less board real estate behind the faceplate and delivers higher linear signal density along the PCB edge. Presently RCx continues to be an advanced solution to meet complex intra rack requirements.

Amphenol ICC has always played its part in providing solutions to meet the changing technologies and challenging demands of the data communications industry. We will continue to stay a step ahead, investing continuously in research and development, making us the preferred choice of our customers.