Memory Interfaces Icon

Memory + Interfaces

28 Gbps Multi-protocol SerDes PHY

The 28 Gbps Multi-protocol SerDes PHYs, including recently acquired Snowbush IP, is a comprehensive 100 Gigabit Ethernet solution that is optimized for power and area in long-reach channels typical of networking and data center applications.

How 28G works

The Rambus 28 Gbps Multi-Protocol SerDes (MPS) PHYs are comprehensive IP solutions that deliver enterprise-class performance across the challenging signaling environments typical of networking and server systems. With high performance and multi-protocol compatibility, the PHYs support data rates from 1.25Gbps to 28.1Gbps across copper and backplane channels with more than 30dB channel insertion loss in a wide range of industry-standard interconnect protocols. They feature application-specific optimization enabled by an efficient and scalable architecture with adaptive and programmable receive equalization, and support for transmit FIR adaptation.

MPSL Subsystem Example

MPSL Subsystem Example

The 28G MPS are designed with a system-oriented approach, taking the interface, interconnect and channel into account when optimizing performance and features to maximize flexibility in today’s most challenging system environments and applications. This makes the PHYs ideal for many long-reach, copper and backplane enterprise environments.

Solution Offerings

  • Multi-protocol PHY supporting data rate in the range of 1.25 Gbps – 28.1 Gbps
  • Optimized for low-power operation and north/south die-edge placement
  • AC-coupled RX front end with on-chip capacitors
  • Flexible ASIC interface for sharing impedance codes among multiple PMA hard macros and reducing the number of external reference resistors for on-chip impedance calibration
  • Duplex Lane configurations of x4 and x1
  • An LC-PLL provides a wide range of operating frequencies
  • Wide range programmable multipliers for reference clock multiplication
  • Differential reference clock inputs are selectively sourced from C4 or internal ASIC interface pins
  • Flexible ASIC clocking
  • Programmable clock outputs from the PLL to the ASIC core
  • Parallel data, transmitted to the PMA, is synchronous to the transmit parallel data clock from the ASIC core
  • Tight lane skew control in the PMA
  • Adaptive receive equalizer with programmable settings providing up to 12dB of CTLE and 8-tap DFE support
  • Support transmit FIR adaptation through back channel for Ethernet applications
  • Embedded microcontroller for improved system configuration loads
  • Built-in Self Test (BIST) support
  • At-speed functional test capability with low-speed reference clocks
  • Built-in PRBS15/31 and custom pattern generation and checking for standalone testing
  • Internal serial loopback with optional phase advancing
  • Parallel loopback supported within the PMA
  • In-situ real-time monitoring and receive data eye schmoo through an adaptive receive sampler
  • ATPG Mux Scan support for digital logic
  • IEEE 1149.6 JTAG boundary scan for serial link pins
  • Multiple interfaces for the access of PMA registers
  • Provides direct register control of all PMA functionality, as well as extended features
  • Interfaces available for connection to a JTAG TAP controller or through a simple parallel read/write port
  • Optional MDIO interface can be provided as required for Ethernet standard PHYs
  • PHY is spec compliant across a wide operating junction temperature range (-20 °C to 125 °C). PLLs, bias circuits, and data path are functional between -40 °C to -20 °C
  • A complete design kit for fast SoC integration
  • PMA Hard Macro
    • Verilog models
    • LEF abstracts (.lef)
    • Timing models (.lib)
    • CDL netlists (.cdl)
    • ATPG models
    • IBIS-AMI models
    • GDSII layout
    • DRC & LVS reports
  • Datasheet
  • SoC Integration guide
  • Optional design integration and bring-up support services

Protocol Compatibility

ProtocolData Rate (Gbps) Application
100GBASE-KR425.78125Copper backplane networking
10GBase-KR10.3125Copper backplane networking
CEI-28G-VSR19.6-28.1Telecom and networking
CEI-28G-SR19.9-28.05Telecom and networking
CEI-28G-MR19.9-28.1Telecom and networking
CEI-25G-LR19.9-25.8Telecom and networking
CEI6-SR4.976-6.375Telecom and networking
CEI11-SR, LR9.95-11.2Telecom and networking
XAUI/2xXAUI3.125, 6.25Chip-to-Chip connectivity
Interlaken 6 G4.976-6.375Networking
JESD 204B/C3.125 – 25High-speed data converters connectivity
CAUI (10x)
XLAUI (4x)
10.3125Chip-to-Chip connectivity
CAUI-428.1Chip-to-Chip connectivity
SFI / XFI10.3125XFP and SFP+ Optical Modules
CPRI1.22 – 24.33Wireless Base Stations
Infiniband EDR25.78125High-performance server connectivity
Infiniband FDR14.0625High-performance server connectivity
SerDes Signal Integrity Challenges at 28Gbps and Beyond

SerDes Signal Integrity Challenges at 28Gbps and Beyond

Maintaining signal integrity has become increasingly difficult as data rates moves past 28Gbps to 56Gbps and beyond. Up to 28Gbps rates, NRZ is the preferred and standardized encoding scheme which consists of 1’s and 0’s. NRZ is also referred to as PAM2 (pulse amplitude modulation, 2-level), due to its two amplitude levels which contain 1 bit of information in every symbol. With serial data rates hitting 56 Gb/s per channel, signal impairments caused by increased bandwidth has prompted the high-speed serial data industry to adopt PAM4, or 4-level pulse amplitude modulation. For PAM4 signals, the baud rate equals one-half the bit rate and the Nyquist frequency equals one-fourth the bit rate. Compared to PAM2/NRZ, PAM4 cuts the bandwidth for a given data rate in half by transmitting two bits in each symbol. This allows engineers to double the bit rate in the channel without doubling the required bandwidth.
Download White Paper

Inventions

Phase Interpolator-Based CDR

Reduces cost, power and area of clock and data recovery (CDR) circuit and improves jitter performance in high-speed parallel and serial links versus PLL-based CDRs.

Related Markets & Applications