As data centers are built and reimagined to support higher-bandwidth computing, storage and content delivery services, the components used to connect them will also need to change – and edge-mount pluggable modules are a good example.
What are Pluggable Modules?
Developed to support telecom and data communications applications, edge-mount pluggable cages on network equipment provide standard interfaces so that pluggable modules can be inserted and adapted.
The pluggable modules can be active (consume power) or passive (do not consume power), and the transmission media can be copper or optical fiber, following industrial networking standards (e.g. Ethernet, Fibre Channel, InfiniBand).
10G SFP+ (one electrical lane) and 40G QSFP+ (four parallel electrical lanes) pluggable modules have been a great market success for years. The same mechanical form factors (SFP28 and 100G QSFP28) have been adopted to support 25G single lane and 4× 25G parallel lanes.
Different physical media can support cost-effective, optimized interconnects for different reaches and data center architecture:
Note: x stands for the number of electrical parallel lanes in the PMD (physical media dependent) sublayer.
New Pluggable Form-Factors and Faster Ethernet Speeds
At 200G and 400G, however, the data transmission technology development (NRZ modulation format) and conventional one-lane SFP and four-lane QSFP form factors can no longer keep pace with imminently high bandwidth requirements.
PAM4 modulation format has been adopted for 50G (and beyond) serial electrical and optical links to combat serious signal integrity degradation at higher frequencies.
New eight-lane form factors (QSFP-DD and OSFP) have been invented to support 8× 25G and 8× 50G PAM4 signaling to support 200G and 400G ports.
QSFP-DD supports the same faceplate density and is backward compatible with QSFP modules, with a maxium power rating of 14W.
OSFP supports eight-lane parallel data transmission with efficient heat dissipation for a wide range of interconnect applications, including metro data center interconnects, with a maxium power rating of 16W.
Smaller-sized, higher-density pluggable form-factors, such as µQSFP and SFP-DD, have been invented to increase switch faceplate I/O density. Cost-optimized form-factors, such as RCx, are available to simplify the pluggable design for high-volume, intra-rack passive copper connectivity.
A new form factor – RCx (RC1, RC2, RC4) – has been introduced to the market as a low-cost passive copper connector and cable system, specifically designed for intra-rack connectivity (<3 m) in 25G, 50G and 100G Ethernet.
This new design has several advantages over the most popular SFP and QSFP pluggables for very large-volume, intra-rack, server-to-switch interconnects:
Make sure to subscribe to our blog updates to find out what else will change as we keep up with the road to 800G Ethernet. We’ll be covering topics like (and click here if you missed our first blog on the topic):
Belden’s complete line of copper and fiber cabling and connectivity solutions will help you get your data center for faster Ethernet speeds. Learn more here. And don’t forget to download our new application note about getting ready for new fiber infrastructure in your data center!
With 13 years of experience in optical communications and photonics device design, Qing Xu is a subject-matter expert in not only optical fiber technology, but also signal transmission, data center trends, fiber/copper connectivity and structured cabling. Joining Belden in 2014, he closely monitors and participates in industry activities related to optical fiber communications systems, data center technology and trends.