QSFP-DD pluggable modules boost data density
With datacenters undergoing rapid growth and 5G development well underway, rising demand and data volume are pushing density and power limits in servers, switches, and other network equipment. Datacenters must be able to support faster processing, more bandwidth, and increased density without sacrificing reliability. Network components can't introduce bandwidth bottlenecks, nor can their temperature exceed design limits.
The QSFP-DD (double density) connector technology will address the density, bandwidth, and thermal issues that arise from bandwidth demands coming first from mobile and IoT devices. Double density refers to the doubling of the number of high-speed electrical interfaces that the QSFP-DD module supports compared with a standard QSFP28 module.
High-speed, high-density pluggable I/O modules provide a highly scalable upgrade path. It all started when SFP pluggable modules ran at 1 Gbps. Then, QSFP optical transceivers added new configuration options and pushed density to aggregate speeds up to 16 Gbps, with four 4 Gbps lanes. QSFP+ takes this to 40 Gbps using four 10 Gbps lanes. Specifically designed for high-density data communications, widely used QSFP+ hot-pluggable transceivers integrate four transmit and four receive channels for greater port density and cost savings over buying higher quantities of traditional SFP+ products.
Next came QSFP28, with increased data rates ranging up to 28 Gbps on each of four lanes, making it popular in datacenter computing, switches, routers, and storage. Supporting next-generation 100 Gbps Ethernet and 100 Gbps InfiniBand Enhanced Data Rate applications, the QSFP28 interconnect transmits up to 25 Gbps per-serial lane data rates with excellent signal integrity, EMI protection, and thermal cooling. Now, double-density SFP (SFP-DD) is in the works for 400 Gbps Ethernet.
QSFP-DD connectors quadruple speeds
With 400 Gbps Ethernet, connectors can't become bottlenecks. The missing link is a connector form factor that provides adequate density to support 256 differential lanes in a 1 RU box while managing thermal and signal integrity.
To achieve 400 Gbps, the interconnect industry is developing next-generation Quad Small Form Factor Pluggable Double Density (QSFP-DD) modules. In 2016, the QSFP-DD multi-source agreement (MSA) was formed to address the technical challenges of achieving a double-density interface that ensures mechanical, electrical, thermal, and signal-integrity interoperability for module components produced by different manufacturers. The group recently released specifications for the new form factor that overcome previous technical issues of specifying a QSFP28 compatible double-density interface. Double density refers to the doubling of the number of high-speed electrical interfaces that the QSFP-DD module supports compared with a QSFP28 module, and the individual lane rate doubles by using Four-Level Pulse Amplitude Modulation (PAM4).
The QSFP-DD specification defines a module, a stacked integrated cage/connector system, and a surface mount cage/connector system (Figure 1). The new form factor expands the standard QSFP four-lane interface by adding a row of contacts providing for an eight-lane electrical interface, each operating up to 25 Gbps with Non-Return-to-Zero (NRZ) modulation or 50 Gbps with PAM4.
Figure 1 A QSFP-DD cage can receive two optical or electrical cable assemblies.
QSFP-DD enables data rates up to 200 Gbps or 400 Gbps aggregate per port. A single switch slot can support up to 36 QSFP-DD modules—providing up to 14.4 Tbps aggregate capacity.