GTS Ethernet Intel® FPGA Hard IP User Guide

Download PDF
ID 817676
Date 4/01/2024
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A newer version of this document is available.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. Overview 2. Getting Started 3. GTS Ethernet Intel® FPGA Hard IP Parameters 4. Integrate GTS Ethernet Intel® FPGA Hard IP to the User Application 5. Functional Description 6. Configuration Registers 7. Document Revision History for the GTS Ethernet Intel® FPGA Hard IP User Guide
1. Overview x
1.1. Agilex™ 5 Ethernet Portfolio and Target Applications 1.2. Agilex™ 5 Ethernet Hard IP Features 1.3. High-Level Functional Overview 1.4. GTS Ethernet Intel® FPGA Hard IP Design Flow
1.2. Agilex™ 5 Ethernet Hard IP Features x
1.2.1. Device Family Support 1.2.2. Device Speed Grade Support 1.2.3. Resource Utilization 1.2.4. Round-Trip Latency 1.2.5. Release Information
2. Getting Started x
2.1. Installing and Licensing GTS Ethernet Intel® FPGA Hard IP Cores 2.2. Generate HDL for Synthesis and Simulation 2.3. Generated File Structure
4. Integrate GTS Ethernet Intel® FPGA Hard IP to the User Application x
4.1. Clocks 4.2. Resets 4.3. MAC Client Interface – Avalon® Streaming Interface 4.4. PCS Client Interface – MII PCS Only 4.5. PCS66 Client Interface – FlexE and OTN 4.6. Connect the Status Interface 4.7. Ethernet Hard IP Reconfiguration Interface 4.8. Precision Time Protocol Interface
4.1. Clocks x
4.1.1. MAC Synchronous Clock Connections to Single Instance 4.1.2. MAC Synchronous Clock Connections to Multiple Instances 4.1.3. Clock Connections to MAC Asynchronous Operation 4.1.4. Clock Connections in PTP-Based Synchronous Operation 4.1.5. Clock Connections in Synchronous Ethernet Operation (Sync-E) 4.1.6. I/O PLL as System PLL
4.2. Resets x
4.2.1. Reset Signals 4.2.2. Reset Sequence 4.2.3. Power Up Reset Sequence 4.2.4. TX Reset Entry Sequence 4.2.5. TX Reset Exit Sequence 4.2.6. RX Reset Entry Sequence 4.2.7. RX Reset Exit Sequence 4.2.8. Auto Recovery Reset Sequence 4.2.9. GTS Reset Sequencer Intel® FPGA IP
4.2.9. GTS Reset Sequencer Intel® FPGA IP x
4.2.9.1. Requirements and Considerations for GTS Reset Sequencer Intel® FPGA IP
4.3. MAC Client Interface – Avalon® Streaming Interface x
4.3.1. TX MAC Avalon Streaming Client Interface 4.3.2. RX MAC Avalon Streaming Client Interface 4.3.3. MAC Flow Control Interface
4.3.1. TX MAC Avalon Streaming Client Interface x
4.3.1.1. TX MAC Avalon Streaming Client Interface with Disabled Preamble Passthrough 4.3.1.2. TX MAC Avalon Streaming Client Interface with Enabled Preamble Passthrough 4.3.1.3. Control Source Address, PAD, and CRC Insertion 4.3.1.4. Assert the i_tx_error to Invalidate a Packet
4.3.2. RX MAC Avalon Streaming Client Interface x
4.3.2.1. RX MAC Avalon Streaming Client Interface with Disabled Preamble Passthrough 4.3.2.2. Enable Preamble Passthrough and RX CRC Forwarding 4.3.2.3. Monitor Status and Errors on the RX MAC Avalon Streaming Client Interface
4.3.3. MAC Flow Control Interface x
4.3.3.1. TX PAUSE and TX PFC Ports 4.3.3.2. Receiving PAUSE and PFC Frames
4.4. PCS Client Interface – MII PCS Only x
4.4.1. PCS Mode TX Interface 4.4.2. PCS Mode RX Interface
4.5. PCS66 Client Interface – FlexE and OTN x
4.5.1. FlexE and OTN Mode TX Interface 4.5.2. FlexE and OTN Mode RX Interface
4.6. Connect the Status Interface x
4.6.1. Status Interface
4.8. Precision Time Protocol Interface x
4.8.1. Time-of-Day Interface 4.8.2. TX 2-Step Timestamp Interface 4.8.3. TX 1-Step Timestamp Interface 4.8.4. RX Timestamp Interface 4.8.5. PTP Status Interface
4.8.3. TX 1-Step Timestamp Interface x
4.8.3.1. PTP Field Offset for 1-Step Operation
5. Functional Description x
5.1. Datapath Description 5.2. GTS Ethernet Intel® FPGA Hard IP MAC 5.3. PCS, OTN, and FlexE Modes 5.4. Precision Time Protocol Interface
5.2. GTS Ethernet Intel® FPGA Hard IP MAC x
5.2.1. MAC TX Datapath 5.2.2. MAC RX Datapath 5.2.3. Congestion and Flow Control Using PAUSE or Priority Flow Control (PFC) 5.2.4. Link Fault Signaling 5.2.5. Order of Ethernet Transmission:
5.2.1. MAC TX Datapath x
5.2.1.1. TX Preamble, Start, and SFD Insertion 5.2.1.2. Source Address Insertion 5.2.1.3. Length/Type Field Processing 5.2.1.4. Frame Padding 5.2.1.5. Frame Check Sequence (CRC-32) Insertion 5.2.1.6. Inter-Packet Gap Generation and Insertion
5.2.2. MAC RX Datapath x
5.2.2.1. RX Preamble Processing 5.2.2.2. RX Strict SFD Checking 5.2.2.3. RX FCS Checking 5.2.2.4. RX Malformed Packet Handling 5.2.2.5. Removing PAD Bytes and FCS Bytes from RX Frames 5.2.2.6. RX Undersized Frames, Oversized Frames, and Frames with Length Errors 5.2.2.7. Inter-Packet Gap
5.2.3. Congestion and Flow Control Using PAUSE or Priority Flow Control (PFC) x
5.2.3.1. Conditions Triggering XOFF Frame Transmission 5.2.3.2. Conditions Triggering XON Frame Transmission 5.2.3.3. Pause Control and Generation Interface 5.2.3.4. Pause Control Frame Filtering
5.3. PCS, OTN, and FlexE Modes x
5.3.1. PCS Mode 5.3.2. OTN Mode 5.3.3. FlexE Mode
5.4. Precision Time Protocol Interface x
5.4.1. Features 5.4.2. PTP User Flow 5.4.3. UI Adjustment 5.4.4. Reference Time Interval 5.4.5. Minimum and Maximum Reference Time (TAM) Interval for UI Measurement (Hardware)
5.4.2. PTP User Flow x
5.4.2.1. PTP TX User Flow 5.4.2.2. PTP RX User Flow
5.4.3. UI Adjustment x
5.4.3.1. TX UI Adjustment 5.4.3.2. RX UI Adjustment
6. Configuration Registers x
6.1. Ethernet Avalon® Memory-Mapped Interface Address Space 6.2. GTS Ethernet Intel® FPGA Hard IP Core CSRs 6.3. Transceiver Avalon® Memory-Mapped Interface Address Space
1. Overview
2. Getting Started
3. GTS Ethernet Intel® FPGA Hard IP Parameters
4. Integrate GTS Ethernet Intel® FPGA Hard IP to the User Application
5. Functional Description
6. Configuration Registers
7. Document Revision History for the GTS Ethernet Intel® FPGA Hard IP User Guide

4.1.5. Clock Connections in Synchronous Ethernet Operation (Sync-E)

When you enable the Synchronous Ethernet (SyncE) operation, two or more channels can share the off-chip cleanup PLL clock output. The Synchronous Ethernet standard, described in the ITU-T G.8261, G.8262, and G.8264 recommendations, requires that the TX clock be filtered to maintain synchronization with the RX reference clock through a sequence of nodes. The expected usage is that user logic drives the transceiver reference clocks with a filtered version of the RX recovered clock signal, to ensure the receive and transmit functions remain synchronized. In this scenario, a design component outside the IP performs the filtering process.

The following figure shows an alternate clocking arrangement for the transceiver reference clock that can be used to enable SyncE operation. the recovered clock outputs o_clk_rec_div or o_clk_rec_div64 or o_cdr_divclk from the GTS Ethernet Intel® FPGA Hard IP can be connected to the off-chip cleanup PLL using REFCLK_GTS pin. In GTS, the dedicated clock output (o_cdr_divclk) from the PMA can be driven to local reference clock pin or dedicated clock output pins if the device has single transceiver bank, whereas the clocks o_clk_rec_div64 and o_clk_rec_div are available at GPIO pins.

For more information about the supported pins, refer Agilex™ 5 Device Family Pin Connection Guidelines.
Figure 14. Clock Connection of Sync-E Clock Through CDR Clock Out Pin
Figure 15. Enable SyncE Option in IP Parameter Editor

Two or more channels share the clock output of an Off-chip Cleanup PLL that meets the specification for a SyncE link. The FPGA provides a primary SyncE clock and a backup SyncE clock to the cleanup PLL. The primary and backup cleanup clocks come from recovered clock output pins from a pair of channels that are both connected to remote stations connected to the same SyncE network, with the transceiver reference clock sourced from the output of the cleanup PLL.

The output frequency is equal to the nominal incoming refclk divided by any predivider on the RX path. The following table shows the recovered clock frequencies with respect to input reference clock.

Table 16.  Recovered Clock Frequency
i_clk_ref((MHz) N Divider o_cdr_divclk(MHz)
156.25 4 39.0625
312.5 4 78.125
322.265625 4 80.5664
Note: If the EHIP System clock is derived from a different reference clock than the transceiver, then the IP must be set to “Custom Cadence” mode to match the PPM difference between the clocks. For SyncE applications, local oscillator should match with recovered clock within +/-4.6ppm.
Level Two Title