Intel® Agilex™ General-Purpose I/O and LVDS SERDES User Guide

Download PDF
ID 683780
Date 10/29/2021
Public

A newer version of this document is available. Customers should click here to go to the newest version.

Document Table of Contents
Document Table of Contents x
1. Intel® Agilex™ General-Purpose I/O and LVDS SERDES Overview 2. Intel® Agilex™ I/O Features and Usage 3. Intel® Agilex™ I/O Termination 4. Intel® Agilex™ High-Speed SERDES I/O Architecture 5. I/O and LVDS SERDES Design Guidelines 6. Troubleshooting Guidelines 7. Documentation Related to the Intel® Agilex™ General-Purpose I/O and LVDS SERDES User Guide 8. Intel® Agilex™ General-Purpose I/O and LVDS SERDES User Guide Archives 9. Document Revision History for the Intel® Agilex™ General-Purpose I/O and LVDS SERDES User Guide
1. Intel® Agilex™ General-Purpose I/O and LVDS SERDES Overview x
1.1. Intel® Agilex™ I/O and Differential I/O Buffers 1.2. Package Selection and I/O Vertical Migration Support 1.3. I/O Banks
1.3. I/O Banks x
1.3.1. I/O Buffer and Registers in Intel® Agilex™ Devices
2. Intel® Agilex™ I/O Features and Usage x
2.1. GPIO Features 2.2. Programmable I/O Element (IOE) Features in Intel® Agilex™ Devices 2.3. I/O Standards and Features for Intel® Agilex™ Configuration Pins 2.4. I/O Implementation Guide
2.1. GPIO Features x
2.1.1. Supported I/O Standards 2.1.2. Intel® Agilex™ I/O Buffer Behavior
2.2. Programmable I/O Element (IOE) Features in Intel® Agilex™ Devices x
2.2.1. Programmable Output Slew Rate Control 2.2.2. Programmable IOE Delay 2.2.3. Programmable Open-Drain Output 2.2.4. Programmable Bus-Hold 2.2.5. Programmable Pull-Up Resistor 2.2.6. Programmable Pre-emphasis 2.2.7. Programmable De-emphasis 2.2.8. Programmable Differential Output Voltage 2.2.9. Schmitt Trigger Input Buffer
2.4. I/O Implementation Guide x
2.4.1. Configuring I/O Assignments in Intel® Quartus® Prime Software 2.4.2. Using GPIO Intel FPGA IP for I/O Implementation
2.4.1. Configuring I/O Assignments in Intel® Quartus® Prime Software x
2.4.1.1. Configuring I/O Assignments Using Assignment Editor 2.4.1.2. Configuring I/O Standards Using Pin Planner
2.4.2. Using GPIO Intel FPGA IP for I/O Implementation x
2.4.2.1. Release Information 2.4.2.2. GPIO Intel FPGA IP Quick Start Guide 2.4.2.3. GPIO Intel® FPGA IP Architecture 2.4.2.4. Verifying Resource Utilization and Design Performance 2.4.2.5. GPIO Intel® FPGA IP Timing 2.4.2.6. GPIO Intel® FPGA IP Design Examples
2.4.2.2. GPIO Intel FPGA IP Quick Start Guide x
2.4.2.2.1. Generating the GPIO Intel FPGA IP ( Intel® Quartus® Prime Pro Edition) 2.4.2.2.2. GPIO Intel® FPGA IP Parameter Settings 2.4.2.2.3. IP Core Generation Output ( Intel® Quartus® Prime Pro Edition) 2.4.2.2.4. GPIO Intel® FPGA IP Interface Signals
2.4.2.3. GPIO Intel® FPGA IP Architecture x
2.4.2.3.1. GPIO Intel® FPGA IP Data Paths 2.4.2.3.2. Register Packing
2.4.2.5. GPIO Intel® FPGA IP Timing x
2.4.2.5.1. Timing Components 2.4.2.5.2. Delay Elements 2.4.2.5.3. Full-Rate or Half-Rate DDIO Output Register 2.4.2.5.4. Timing Analysis 2.4.2.5.5. Timing Closure Guidelines
2.4.2.6. GPIO Intel® FPGA IP Design Examples x
2.4.2.6.1. GPIO Intel FPGA IP Synthesizable Intel® Quartus® Prime Design Example 2.4.2.6.2. GPIO Intel FPGA IP Simulation Design Example
3. Intel® Agilex™ I/O Termination x
3.1. Single-Ended I/O Termination in Intel® Agilex™ Devices 3.2. True Differential Signaling I/O Termination
3.1. Single-Ended I/O Termination in Intel® Agilex™ Devices x
3.1.1. Single-Ended I/O Standard OCT Termination 3.1.2. OCT Calibration Block 3.1.3. Single-ended I/O Standards External Termination 3.1.4. Single-ended I/O Termination Implementation Guide
3.1.1. Single-Ended I/O Standard OCT Termination x
3.1.1.1. RS OCT 3.1.1.2. RT OCT 3.1.1.3. Dynamic OCT
3.1.4. Single-ended I/O Termination Implementation Guide x
3.1.4.1. Configuring OCT Using Assignment Editor 3.1.4.2. Using OCT Intel FPGA IP for I/O Termination Implementation
3.1.4.2. Using OCT Intel FPGA IP for I/O Termination Implementation x
3.1.4.2.1. Release Information 3.1.4.2.2. OCT Intel® FPGA IP Quick Start Guide 3.1.4.2.3. OCT Intel® FPGA IP Architecture 3.1.4.2.4. OCT Intel® FPGA IP Design Example
3.2. True Differential Signaling I/O Termination x
3.2.1. External I/O Termination 3.2.2. True Differential Signaling I/O Standard OCT Termination
3.2.2. True Differential Signaling I/O Standard OCT Termination x
3.2.2.1. Configuring Differential Input RD OCT Using Assignment Editor 3.2.2.2. Differential Input RD OCT Restrictions and Guidelines
4. Intel® Agilex™ High-Speed SERDES I/O Architecture x
4.1. Intel® Agilex™ High-Speed SERDES I/O Overview 4.2. Using LVDS SERDES Intel FPGA IP for High-Speed LVDS I/O Implementation 4.3. Intel® Agilex™ LVDS SERDES Transmitter 4.4. Intel® Agilex™ LVDS SERDES Receiver 4.5. Intel® Agilex™ LVDS Interface with External PLL Mode 4.6. LVDS SERDES IP Initialization and Reset 4.7. Intel® Agilex™ LVDS SERDES Source-Synchronous Timing Budget 4.8. LVDS SERDES IP Timing 4.9. LVDS SERDES IP Design Examples
4.1. Intel® Agilex™ High-Speed SERDES I/O Overview x
4.1.1. High-Speed SERDES Architecture 4.1.2. Intel® Agilex™ GPIO Banks, SERDES, and DPA Locations
4.2. Using LVDS SERDES Intel FPGA IP for High-Speed LVDS I/O Implementation x
4.2.1. Release Information 4.2.2. LVDS SERDES Intel FPGA IP Quick Start Guide
4.2.2. LVDS SERDES Intel FPGA IP Quick Start Guide x
4.2.2.1. Generating the LVDS SERDES Intel FPGA IP ( Intel® Quartus® Prime Pro Edition) 4.2.2.2. LVDS SERDES Intel® FPGA IP Parameter Settings 4.2.2.3. IP Generation Output ( Intel® Quartus® Prime Pro Edition) 4.2.2.4. LVDS SERDES IP Signals
4.2.2.2. LVDS SERDES Intel® FPGA IP Parameter Settings x
4.2.2.2.1. LVDS SERDES IP General Settings 4.2.2.2.2. LVDS SERDES IP PLL Settings 4.2.2.2.3. LVDS SERDES IP Receiver Settings 4.2.2.2.4. LVDS SERDES IP Transmitter Settings 4.2.2.2.5. LVDS SERDES IP Clock Resource Summary
4.3. Intel® Agilex™ LVDS SERDES Transmitter x
4.3.1. LVDS SERDES Transmitter Blocks 4.3.2. Serializer Bypass for DDR and SDR Operations 4.3.3. Serializer 4.3.4. Differential I/O Bit Position 4.3.5. Clocking Differential Transmitters
4.3.5. Clocking Differential Transmitters x
4.3.5.1. Setting the Transmitter Output Clock Parameters
4.4. Intel® Agilex™ LVDS SERDES Receiver x
4.4.1. LVDS SERDES Receiver Blocks 4.4.2. Clocking LVDS SERDES Receivers 4.4.3. LVDS SERDES Receiver Modes
4.4.1. LVDS SERDES Receiver Blocks x
4.4.1.1. Receiver Blocks in Intel® Agilex™ Devices
4.4.1.1. Receiver Blocks in Intel® Agilex™ Devices x
4.4.1.1.1. DPA Block 4.4.1.1.2. Synchronizer (DPA FIFO) 4.4.1.1.3. Data Realignment Block (Bit Slip) 4.4.1.1.4. Deserializer
4.4.2. Clocking LVDS SERDES Receivers x
4.4.2.1. Receiver Input Clock Parameters Setup
4.4.3. LVDS SERDES Receiver Modes x
4.4.3.1. Non-DPA Mode 4.4.3.2. DPA Mode 4.4.3.3. Soft-CDR Mode
4.5. Intel® Agilex™ LVDS Interface with External PLL Mode x
4.5.1. IOPLL IP Signal Interface with LVDS SERDES IP 4.5.2. IOPLL Parameter Values for External PLL Mode 4.5.3. Connection between IOPLL IP and LVDS SERDES IP in External PLL Mode
4.6. LVDS SERDES IP Initialization and Reset x
4.6.1. Initializing the LVDS SERDES IP in Non-DPA Mode 4.6.2. Initializing the LVDS SERDES IP in DPA Mode 4.6.3. Resetting the DPA 4.6.4. Word Boundaries Alignment
4.6.4. Word Boundaries Alignment x
4.6.4.1. Aligning Word Boundaries
4.7. Intel® Agilex™ LVDS SERDES Source-Synchronous Timing Budget x
4.7.1. Transmitter Channel-to-Channel Skew 4.7.2. Receiver Skew Margin
4.8. LVDS SERDES IP Timing x
4.8.1. I/O Timing Analysis 4.8.2. FPGA Timing Analysis 4.8.3. Timing Analysis for the External PLL Mode
4.8.1. I/O Timing Analysis x
4.8.1.1. Obtaining RSKM Report 4.8.1.2. Obtaining TCCS Report
4.9. LVDS SERDES IP Design Examples x
4.9.1. LVDS SERDES IP Synthesizable Intel® Quartus® Prime Design Examples 4.9.2. LVDS SERDES IP Simulation Design Example 4.9.3. Combined LVDS SERDES IP Transmitter and Receiver Design Example 4.9.4. LVDS SERDES IP Dynamic Phase Shift Design Example
5. I/O and LVDS SERDES Design Guidelines x
5.1. Intel® Agilex™ I/O Design Guidelines 5.2. Intel® Agilex™ LVDS SERDES Design Guidelines
5.1. Intel® Agilex™ I/O Design Guidelines x
5.1.1. VREF Sources and VREF Pins 5.1.2. I/O Standards Implementation based on VCCIO_PIO Voltages 5.1.3. OCT Calibration Block Requirement 5.1.4. Placement Requirements 5.1.5. Simultaneous Switching Noise (SSN) 5.1.6. Special Pins Requirement 5.1.7. External Memory Interface Pin Placement Requirements 5.1.8. HPS Shared I/O Requirements 5.1.9. Clocking Requirements 5.1.10. SDM Shared I/O Requirements 5.1.11. Configuration Pins 5.1.12. Unused Pins 5.1.13. Voltage Setting for Unused I/O Banks 5.1.14. Guidelines for I/O Pins in GPIO, HPS, and SDM Banks During Power Sequencing 5.1.15. Drive Strength Requirement for GPIO Input Pins 5.1.16. Maximum DC Current Restrictions 5.1.17. 1.2 V I/O Interface Voltage Level Compatibility 5.1.18. GPIO Pins for Avalon-ST Configuration Scheme 5.1.19. Maximum True Differential Signaling RX Pairs Per I/O Lane 5.1.20. I/O Simulation
5.1.20. I/O Simulation x
5.1.20.1. HSPICE* Models 5.1.20.2. Net Length Reports
5.2. Intel® Agilex™ LVDS SERDES Design Guidelines x
5.2.1. Use PLLs in Integer PLL Mode for LVDS 5.2.2. Use High-Speed Clock from PLL to Clock SERDES Only 5.2.3. Pin Placement for Differential Channels 5.2.4. SERDES Pin Pairs for Soft-CDR Mode 5.2.5. Placing LVDS Transmitters and Receivers in the Same I/O Bank
5.2.5. Placing LVDS Transmitters and Receivers in the Same I/O Bank x
5.2.5.1. Using an External PLL
1. Intel® Agilex™ General-Purpose I/O and LVDS SERDES Overview
2. Intel® Agilex™ I/O Features and Usage
3. Intel® Agilex™ I/O Termination
4. Intel® Agilex™ High-Speed SERDES I/O Architecture
5. I/O and LVDS SERDES Design Guidelines
6. Troubleshooting Guidelines
7. Documentation Related to the Intel® Agilex™ General-Purpose I/O and LVDS SERDES User Guide
8. Intel® Agilex™ General-Purpose I/O and LVDS SERDES User Guide Archives
9. Document Revision History for the Intel® Agilex™ General-Purpose I/O and LVDS SERDES User Guide

Figure 18.  High-Level View of Single-Ended GPIO


Table 28.   GPIO IP Data Path Modes
Data Path Register Mode
Bypass Simple Register DDR I/O
Full-Rate Half-Rate
Input Data goes from the delay element to the core, bypassing all double data rate I/Os (DDIOs). The full-rate DDIO operates as a simple register, bypassing half-rate DDIOs. The Fitter chooses whether to pack the register in the I/O or implement the register in the core, depending on the area and timing trade-offs. The full-rate DDIO operates as a regular DDIO, bypassing the half-rate DDIOs. The full-rate DDIO operates as a regular DDIO. The half-rate DDIOs convert full-rate data to half-rate data.
Output Data goes from the core straight to the delay element, bypassing all DDIOs. The full-rate DDIO operates as a simple register, bypassing half-rate DDIOs. The Fitter chooses whether to pack the register in the I/O or implement the register in the core, depending on the area and timing trade-offs. The full-rate DDIO operates as a regular DDIO, bypassing the half-rate DDIOs. The full-rate DDIO operates as a regular DDIO. The half-rate DDIOs convert full-rate data to half-rate data.
Bidirectional The output buffer drives both an output pin and an input buffer. The full-rate DDIO operates as a simple register. The output buffer drives both an output pin and an input buffer. The full-rate DDIO operates as a regular DDIO. The output buffer drives both an output pin and an input buffer. The input buffer drives a set of three flip-flops. The full-rate DDIO operates as a regular DDIO. The half-rate DDIOs convert full-rate data to half-rate. The output buffer drives both an output pin and an input buffer. The input buffer drives a set of three flip-flops.

If you use asynchronous clear and preset signals, all DDIOs share these same signals.

Half-rate and full-rate DDIOs connect to separate clocks. When you use half-rate and full-rate DDIOs, the full-rate clock must run at twice the half-rate frequency. You can use different phase relationships to meet timing requirements.

Input Path

The pad sends data to the input buffer, and the input buffer feeds the delay element. After the data goes to the output of the delay element, the programmable bypass multiplexers select the features and paths to use.

Each input path contains two stages of DDIOs, which are full-rate and half-rate.

Figure 19. Simplified View of Single-Ended GPIO Input Path


  1. The pad receives data.
  2. DDIO IN (1) captures data on the rising and falling edges of ck_fr and sends the data, signals (A) and (B) in the following waveform figure, at single data rate.
  3. DDIO IN (2) and DDIO IN (3) halve the data rate.
  4. dout[3:0] presents the data as a half-rate bus.
Figure 20. Input Path Waveform in DDIO Mode with Half-Rate Conversion

In this figure, the data goes from full-rate clock at double data rate to half-rate clock at single data rate. The data rate is divided by four and the bus size is increased by the same ratio. The overall throughput through the GPIO IP remains unchanged.

The actual timing relationship between different signals may vary depending on the specific design, delays, and phases that you choose for the full-rate and half-rate clocks.



Note: The GPIO IP does not support dynamic calibration of bidirectional pins.

Output and Output Enable Paths

The output delay element sends data to the pad through the output buffer.

Each output path contains two stages of DDIOs, which are half-rate and full-rate.

Figure 21. Simplified View of Single-Ended GPIO Output Path


Figure 22. Output Path Waveform in DDIO Mode with Half-Rate Conversion
Figure 23.  Simplified View of Output Enable Path


The difference between the output path and output enable (OE) path is that the OE path does not contain full-rate DDIO. To support packed-register implementations in the OE path, a simple register operates as full-rate DDIO. For the same reason, only one half-rate DDIO is present.

The OE path operates in the following three fundamental modes:

  • Bypass—the core sends data directly to the delay element, bypassing all DDIOs.
  • Packed Register—bypasses half-rate DDIO.
  • SDR output at half-rate—half-rate DDIOs convert data from full-rate to half-rate.
Note: The GPIO IP does not support dynamic calibration of bidirectional pins.
Level Two Title