Interface IP chip icon

Interface IP

DDR4 PHY

Designed to meet the memory-intensive workload demands of networking and data center applications, the DDR4 memory PHY delivers maximum performance and power efficiency while maintaining full compatibility with the DDR4 and DDR3 industry standards. With the Northwest Logic DDR4 Controller it comprises a complete DDR4 memory interface subsystem.

How the DDR4 Interface works

The Rambus DDR4 memory PHY delivers industry-leading data rates of up to 3200 Mbps and is compatible with the DDR4 and DDR3 standards. The PHY consists of a Command/Address (C/A) macro cell and Data (DQ) macro cells configured to create a 72-bit wide channel.

The Northwest Logic DDR4 controller maximizes memory bus efficiency via Look-Ahead command processing, bank management, auto-precharge and additive latency support. The core is DFI compatible and supports a range of interfaces to user logic.

DDR4 Memory Interface Subsystem
DDR4 Memory Interface Subsystem

The Rambus DDR4 PHY and Northwest Logic DDR4 controller used together comprise a complete DDR4 memory interface subsystem. Alternatively, these cores can be licensed separately to be paired with 3rd-party DDR4 controller or PHY solutions.

Solution Offerings

  • Autonomous initialization (PHY independent mode)
  • Support for multiple channel configurations from x16 through x72 per channel
  • Support for single channel, 1up to 4 ranks
  • Selectable low-power operating states
  • DFI 4.0 and 3.1 compliant for easy integration with memory controller
  • Programmable output impedance and on-die termination
  • Available in North-South, East-West, and Corner configurations
  • ZQ calibration of output impedance and on-die calibration
  • Utilizes standard 8-layer 602011- or 13- metal layer stack.
  • Supports package-on-package and C4 flip-chip packaging options
  • Register interface for state observation
  • Test traffic generation and error checking for in-situ test
  • LabStation™ software environment for system level bring-up, characterization, and validation
  • Fully-characterized hard macro (GDSII):
  • North-South side of SOC Placement
  • Complete design views:
    • Gate-level and IO models
    • Verification test benches
    • Layout abstracts (.lef)
    • Timing models (.lib)
  • Memory controller reference design
  • Full Documentation
    • Integration Guidelines
    • Package and PCB design guidelines
    • ASIC/DFT manufacturing guidelines
    • Test and Characterization user guide
    • Verilog models
    • CDL netlists *(.cdl)
    • ATPG models
    • GDSII layout
    • DRC & LVS reports

Comprehensive Chip and System Design Reviews

  • Kickoff/Program Review
  • Floor plan Review
  • Test/Characterization Plan Review
  • Package Design Review
  • Board Design Review
  • Final Chip Integration Review
  • Bring-up and Test Review
 

Engineering Design Services:

  • Package design
  • System board layout
  • Statistically-based signal and power integrity analysis
2.5D/3D Packaging Solutions for AI and HPC

Memory Systems for AI and Leading-Edge Applications

Thanks to rapid advancements in computing, neural networks are fueling tremendous growth in AI for a broad spectrum of applications. Learn about the memory architectures, and their relative advantages, at the heart of the AI revolution.

Protocol Compatibility

ProtocolData Rate (Gbps) Application
DDR41600-3200Data Center and Networking
DDR3 (1.5V)1066-2133Data Center and Networking
DDR3L (1.35V)1066-2133Data Center and Networking
DDR3U (1.25V)1066-2133Data Center and Networking

Inventions

FlexPhase™ Timing Adjustment Circuits

FlexPhase per bit timing adjustment circuits deskew data and clock signals to improve signal integrity and simplify package and PCB system design.

Output Driver Calibration

Output driver calibration allows SoC designers to tune the output signaling to optimal levels in order to improve data rates and system voltage margin.

On Die Termination Calibration

ODT calibration enables higher data rates and superior DRAM device and module performance by establishing the optimal termination value to compensate for variations in process and operating conditions that degrade signal performance.

Upcoming Webinar: AI Requires Tailored DRAM Solutions