Triple-Speed Ethernet IP User Guide: Agilex™ 3 and Agilex™ 5 FPGAs and SoCs

Download PDF
ID 813669
Date 8/04/2025
Public
Document Table of Contents
Document Table of Contents x
1. About Triple-Speed Ethernet IP for Agilex™ 3 and Agilex™ 5 devices 2. Getting Started 3. Parameter Settings 4. Functional Description 5. Configuration Register Space 6. Interface Signals 7. Design Considerations 8. Timing Constraints 9. Testbench 10. Triple-Speed Ethernet Debug Checklist 11. Software Programming Interface 12. Triple-Speed Ethernet IP User Guide: Agilex™ 3 and Agilex™ 5 FPGAs and SoCs Archives 13. Document Revision History for the Triple-Speed Ethernet IP User Guide: Agilex™ 3 and Agilex™ 5 FPGAs and SoCs A. Ethernet Frame Format B. Simulation Parameters
1. About Triple-Speed Ethernet IP for Agilex™ 3 and Agilex™ 5 devices x
1.1. Release Information 1.2. Device Family Support 1.3. Device Speed Grade Support 1.4. Features 1.5. 10/100/1000 Ethernet MAC Versus Small MAC 1.6. High-Level Block Diagrams 1.7. Performance and Resource Utilization
2. Getting Started x
2.1. Introduction to Altera IP Cores 2.2. Installing and Licensing IPs 2.3. Specifying the IP Core Parameters and Options 2.4. IP Core Generation Output 2.5. Simulating IP Cores
2.2. Installing and Licensing IPs x
2.2.1. Intel FPGA IP Evaluation Mode
2.4. IP Core Generation Output x
2.4.1. Design Constraint File
3. Parameter Settings x
3.1. Core Configuration 3.2. Ethernet MAC Options 3.3. FIFO Options 3.4. Timestamp Options 3.5. PCS/Transceiver Options 3.6. Analog Parameter Settings 3.7. LVDS Pin Settings
4. Functional Description x
4.1. 10/100/1000 Ethernet MAC 4.2. 1000BASE-X/SGMII PCS With Optional Embedded PMA (GTS) 4.3. HPS GMII Adapter 4.4. Precision Time Protocol 4.5. Deterministic Latency
4.1. 10/100/1000 Ethernet MAC x
4.1.1. MAC Architecture 4.1.2. MAC Interfaces 4.1.3. MAC Transmit Datapath 4.1.4. MAC Receive Datapath 4.1.5. MAC Transmit and Receive Latencies 4.1.6. FIFO Buffer Thresholds 4.1.7. Congestion and Flow Control 4.1.8. Magic Packets 4.1.9. MAC Local Loopback 4.1.10. MAC Reset 4.1.11. PHY Management (MDIO) 4.1.12. Connecting MAC to External PHYs
4.1.3. MAC Transmit Datapath x
4.1.3.1. IP Payload Re-alignment 4.1.3.2. Address Insertion 4.1.3.3. Frame Payload Padding 4.1.3.4. CRC-32 Generation 4.1.3.5. Interpacket Gap Insertion 4.1.3.6. Collision Detection in Half-Duplex Mode
4.1.4. MAC Receive Datapath x
4.1.4.1. Preamble Processing 4.1.4.2. Collision Detection in Half-Duplex Mode 4.1.4.3. Address Checking 4.1.4.4. Frame Type Validation 4.1.4.5. Payload Pad Removal 4.1.4.6. CRC Checking 4.1.4.7. Length Checking 4.1.4.8. Frame Writing 4.1.4.9. IP Payload Alignment
4.1.4.3. Address Checking x
4.1.4.3.1. Unicast Address Checking 4.1.4.3.2. Multicast Address Resolution
4.1.6. FIFO Buffer Thresholds x
4.1.6.1. Receive Thresholds 4.1.6.2. Transmit Thresholds 4.1.6.3. Transmit FIFO Buffer Underflow
4.1.7. Congestion and Flow Control x
4.1.7.1. Remote Device Congestion 4.1.7.2. Receive FIFO Buffer and Local Device Congestion
4.1.8. Magic Packets x
4.1.8.1. Sleep Mode 4.1.8.2. Magic Packet Detection
4.1.11. PHY Management (MDIO) x
4.1.11.1. MDIO Connection 4.1.11.2. MDIO Frame Format
4.1.12. Connecting MAC to External PHYs x
4.1.12.1. Gigabit Ethernet 4.1.12.2. Programmable 10/100 Ethernet 4.1.12.3. Programmable 10/100/1000 Ethernet Operation
4.2. 1000BASE-X/SGMII PCS With Optional Embedded PMA (GTS) x
4.2.1. 1000BASE-X/SGMII PCS Architecture 4.2.2. Transmit Operation 4.2.3. Receive Operation 4.2.4. Transmit and Receive Latencies 4.2.5. GMII Converter 4.2.6. SGMII Converter 4.2.7. Auto-Negotiation 4.2.8. Ten-bit Interface 4.2.9. 1000BASE-X/SGMII PCS Reset
4.2.2. Transmit Operation x
4.2.2.1. Frame Encapsulation 4.2.2.2. 8b/10b Encoding
4.2.3. Receive Operation x
4.2.3.1. Comma Detection 4.2.3.2. 8b/10b Decoding 4.2.3.3. Frame De-encapsulation 4.2.3.4. Synchronization 4.2.3.5. Carrier Sense 4.2.3.6. Collision Detection
4.2.6. SGMII Converter x
4.2.6.1. Transmit 4.2.6.2. Receive
4.2.7. Auto-Negotiation x
4.2.7.1. 1000BASE-X Auto-Negotiation 4.2.7.2. SGMII Auto-Negotiation
4.4. Precision Time Protocol x
4.4.1. IEEE 1588v2 Supported Configurations 4.4.2. IEEE 1588v2 Features 4.4.3. IEEE 1588v2 Architecture 4.4.4. IEEE 1588v2 Transmit Datapath 4.4.5. IEEE 1588v2 Receive Datapath 4.4.6. IEEE 1588v2 Frame Format
4.4.6. IEEE 1588v2 Frame Format x
4.4.6.1. PTP Frame in IEEE 802.3 4.4.6.2. PTP Frame over UDP/IPv4 4.4.6.3. PTP Frame over UDP/IPv6
5. Configuration Register Space x
5.1. MAC Configuration Register Space 5.2. PCS Configuration Register Space 5.3. Register Initialization
5.1. MAC Configuration Register Space x
5.1.1. Base Configuration Registers (Dword Offset 0x00 – 0x17) 5.1.2. Statistics Counters (Dword Offset 0x18 – 0x38) 5.1.3. Transmit and Receive Command Registers (Dword Offset 0x3A – 0x3B) 5.1.4. Supplementary Address (Dword Offset 0xC0 – 0xC7) 5.1.5. IEEE 1588v2 Feature (Dword Offset 0xD0 – 0xD6) 5.1.6. Deterministic Latency (Dword Offset 0xE1– 0xE3) 5.1.7. IEEE 1588v2 Feature PMA Delay
5.1.1. Base Configuration Registers (Dword Offset 0x00 – 0x17) x
5.1.1.1. Command_Config Register (Dword Offset 0x02)
5.2. PCS Configuration Register Space x
5.2.1. Control Register (Word Offset 0x00) 5.2.2. Status Register (Word Offset 0x01) 5.2.3. Dev_Ability and Partner_Ability Registers (Word Offset 0x04 – 0x05) 5.2.4. An_Expansion Register (Word Offset 0x06) 5.2.5. If_Mode Register (Word Offset 0x14)
5.2.3. Dev_Ability and Partner_Ability Registers (Word Offset 0x04 – 0x05) x
5.2.3.1. 1000BASE-X 5.2.3.2. SGMII MAC Mode Auto Negotiation 5.2.3.3. SGMII PHY Mode Auto Negotiation
5.3. Register Initialization x
5.3.1. Triple-Speed Ethernet System with MII/GMII or RGMII 5.3.2. Triple-Speed Ethernet System with SGMII 5.3.3. Triple-Speed Ethernet System with 1000BASE-X Interface
6. Interface Signals x
6.1. Interface Signals 6.2. Timing
6.1. Interface Signals x
6.1.1. 10/100/1000 Ethernet MAC Signals 6.1.2. 10/100/1000 Multiport Ethernet MAC Signals 6.1.3. 10/100/1000 Ethernet MAC with 1000BASE-X/SGMII PCS Signals 6.1.4. 10/100/1000 Ethernet MAC with Internal FIFO Buffers, and 1000BASE-X/SGMII 2XTBI PCS with Embedded PMA (GTS) Signals 6.1.5. 10/100/1000 Multiport Ethernet MAC with 1000BASE-X/SGMII PCS Signals 6.1.6. 1000BASE-X/SGMII PCS Signals 6.1.7. 1000BASE-X/SGMII PCS and PMA (LVDS) Signals 6.1.8. 1000BASE-X/SGMII 2XTBI PCS Signals 6.1.9. 10/100/1000 Ethernet MAC with 1000BASE-X/SGMII PCS and Embedded PMA (LVDS) Signals 6.1.10. 10/100/1000 Multiport Ethernet MAC with 1000BASE-X/SGMII PCS and Embedded PMA (LVDS) Signals 6.1.11. 10/100/1000 Ethernet MAC without Internal FIFO Buffers with 1000BASE-X/SGMII 2XTBI PCS and Embedded PMA Signals (GTS) with IEEE 1588v2 6.1.12. 10/100/1000 Multiport Ethernet MAC with 1000BASE-X/SGMII PCS and Embedded PMA Signals (LVDS) with IEEE 1588v2
6.1.1. 10/100/1000 Ethernet MAC Signals x
6.1.1.1. Clock and Reset Signals 6.1.1.2. Clock Enabler Signals 6.1.1.3. MAC Control Interface Signals 6.1.1.4. MAC Status Signals 6.1.1.5. MAC Receive Interface Signals 6.1.1.6. MAC Transmit Interface Signals 6.1.1.7. Pause and Magic Packet Signals 6.1.1.8. MII/GMII/RGMII Signals 6.1.1.9. PHY Management Signals
6.1.2. 10/100/1000 Multiport Ethernet MAC Signals x
6.1.2.1. Multiport MAC Clock and Reset Signals 6.1.2.2. Multiport MAC Receive Interface Signals 6.1.2.3. Multiport MAC Transmit Interface Signals 6.1.2.4. Multiport MAC Packet Classification Signals 6.1.2.5. Multiport MAC FIFO Status Signals
6.1.3. 10/100/1000 Ethernet MAC with 1000BASE-X/SGMII PCS Signals x
6.1.3.1. TBI Interface Signals 6.1.3.2. Status LED Control Signals 6.1.3.3. SERDES Control Signals
6.1.4. 10/100/1000 Ethernet MAC with Internal FIFO Buffers, and 1000BASE-X/SGMII 2XTBI PCS with Embedded PMA (GTS) Signals x
6.1.4.1. GMII Clock Signals 6.1.4.2. GTS Transceiver Direct PHY Signals 6.1.4.3. GTS Reset Sequencer Signals 6.1.4.4. PMA Reconfiguration Interface Signals
6.1.6. 1000BASE-X/SGMII PCS Signals x
6.1.6.1. PCS Control Interface Signals 6.1.6.2. PCS Reset Signals 6.1.6.3. MII/GMII Clocks and Clock Enablers 6.1.6.4. GMII 6.1.6.5. MII 6.1.6.6. SGMII Status Signals
6.1.8. 1000BASE-X/SGMII 2XTBI PCS Signals x
6.1.8.1. Clock Enablers 6.1.8.2. PCS Reset Signals 6.1.8.3. TBI Interface Signals
6.1.9. 10/100/1000 Ethernet MAC with 1000BASE-X/SGMII PCS and Embedded PMA (LVDS) Signals x
6.1.9.1. 1.25 Gbps LVDS Serial Interface
6.1.11. 10/100/1000 Ethernet MAC without Internal FIFO Buffers with 1000BASE-X/SGMII 2XTBI PCS and Embedded PMA Signals (GTS) with IEEE 1588v2 x
6.1.11.1. Deterministic Latency Clock Signals 6.1.11.2. IEEE 1588v2 RX Timestamp Signals 6.1.11.3. IEEE 1588v2 TX Timestamp Signals 6.1.11.4. IEEE 1588v2 TX Timestamp Request Signals 6.1.11.5. IEEE 1588v2 TX Insert Control Timestamp Signals 6.1.11.6. IEEE 1588v2 Time-of-Day (TOD) Clock Interface Signals
6.2. Timing x
6.2.1. Avalon Streaming Receive Interface 6.2.2. Avalon Streaming Transmit Interface 6.2.3. GMII Transmit 6.2.4. GMII Receive 6.2.5. RGMII Transmit 6.2.6. RGMII Receive 6.2.7. MII Transmit 6.2.8. MII Receive
7. Design Considerations x
7.1. Exposed Ports in the New User Interface 7.2. Clocking Scheme of MAC with 2XTBI PCS and Embedded PMA (GTS)
8. Timing Constraints x
8.1. Creating Clock Constraints 8.2. Recommended Clock Frequency
9. Testbench x
9.1. Triple-Speed Ethernet Testbench Architecture 9.2. Testbench Components 9.3. Testbench Verification 9.4. Testbench Configuration 9.5. Test Flow 9.6. Simulation Model
9.6. Simulation Model x
9.6.1. Generate the Simulation Model 9.6.2. Simulate the IP 9.6.3. Simulation Model Files
10. Triple-Speed Ethernet Debug Checklist x
10.1. Debug Checklist 10.2. Case Study
10.2. Case Study x
10.2.1. Avalon Streaming Interface 10.2.2. Statistic Counter 10.2.3. Packet Drop 10.2.4. Congestion and Flow Control 10.2.5. Loopback 10.2.6. Link Synchronization
11. Software Programming Interface x
11.1. Driver Architecture 11.2. Directory Structure 11.3. PHY Definition 11.4. Using Multiple SG-DMA Descriptors 11.5. Using Jumbo Frames 11.6. API Functions 11.7. Constants
11.6. API Functions x
11.6.1. alt_tse_mac_get_common_speed() 11.6.2. alt_tse_mac_set_common_speed() 11.6.3. alt_tse_phy_add_profile() 11.6.4. alt_tse_system_add_sys() 11.6.5. tse_mac_setMIImode() 11.6.6. tse_mac_SwReset()
A. Ethernet Frame Format x
A.1. Basic Frame Format A.2. VLAN and Stacked VLAN Frame Format A.3. Pause Frame Format
A.3. Pause Frame Format x
A.3.1. Pause Frame Generation
B. Simulation Parameters x
B.1. Functionality Configuration Parameters B.2. Test Configuration Parameters
1. About Triple-Speed Ethernet IP for Agilex™ 3 and Agilex™ 5 devices
2. Getting Started
3. Parameter Settings
4. Functional Description
5. Configuration Register Space
6. Interface Signals
7. Design Considerations
8. Timing Constraints
9. Testbench
10. Triple-Speed Ethernet Debug Checklist
11. Software Programming Interface
12. Triple-Speed Ethernet IP User Guide: Agilex™ 3 and Agilex™ 5 FPGAs and SoCs Archives
13. Document Revision History for the Triple-Speed Ethernet IP User Guide: Agilex™ 3 and Agilex™ 5 FPGAs and SoCs
A. Ethernet Frame Format
B. Simulation Parameters

4.5. Deterministic Latency

The Deterministic Latency (DL) term used across this document refers to the ability to precisely determine the delay between the elastic FIFO (EFIFO) and the PMA pins.

The deterministic latency measurement methodology is based on the concept of measuring the time when a given word is at the interface to the PMA and when that same word is at the FPGA core. The difference in time between these two events, when added to the PMA propagation delay, determines the total latency between the FPGA core and the serial pins. Such a calculation intrinsically includes all delays due to intermediate logic, FIFOs, and all other effects.

You must turn on Enable timestamping option to enable the DL feature.
Note: Deterministic latency feature is only supported on the 10/100/1000 Ethernet MAC without Internal FIFO Buffers with 1000BASE-X/SGMII 2XTBI PCS and Embedded PMA Signals (GTS) with IEEE 1588v2 variant.
Table 29.  Deterministic Latency Parameter Description for Triple-Speed Ethernet This table shows parameter description for the deterministic latency measurement for Triple-Speed Ethernet use case.
Item Value Description
sampling_clk period 4.375 ns Period for sampling clock (i_dl_sampling_clk) of 228.571429 MHz.
UI period 0.8 ns Period for unit interval.
parallel_clk 20 UI Period for 1 parallel clock cycle.
tx_delay (TxDL) Read from EFIFO-DL register 0xE2[20:0]

TX delay value in sampling_clk cycles, fixed point format Q13.8.

Bit [20:8] is integer, bit [7:0] is fractional number.

For example, tx_delay = 0x27F4,

Bit [20:8] = 0x27 = 39

Bit [7:0] = 0xF4 = 0.953125

Hence, tx_delay = 39.953125 clock cycles.

Or tx_delay = 39.953125 × (4.375ns sampling_clk) = 174.794921875ns
rx_delay (RxDL) Read from EFIFO-DL register 0xE3[20:0]

RX delay value in the sampling_clk cycles, fixed point format Q13.8.

Bit [20:8] is integer, bit [7:0] is fractional number.

For example, rx_delay = 0x27F4,

Bit [20:8] = 0x27 = 39

Bit [7:0] = 0xF4 = 0.953125

Hence, tx_delay = 39.953125 clock cycles.

Or rx_delay = 39.953125 × (4.375ns sampling_clk) = 174.794921875ns
tx_pma_delay

Simulation: 49

Hardware: 49

 TX PMA delay in number of UI. Multiply by UI period to convert to nanosecond and fractional nanosecond format.
rx_pma_delay

Simulation: 67.5

Hardware: 67.5

 RX PMA delay in number of UI. Multiply by UI period to convert to nanosecond and fractional nanosecond format.
Table 30.  Deterministic Latency Measurement for Triple-Speed Ethernet This table shows the TX and RX latency calculations for the 1G.
Variant TX Latency (ns) RX Latency (ns)
1G TxDL * (sampling_clock period in ns)/(2^8) + TX PMA Delay RxDL * (sampling_clock period in ns)/(2^8) + RX PMA Delay
Note:
  1. Read TX/RX DL values from DL soft registers 0xE2 and 0xE3 respectively and calculate TX/RX latency.
  2. Convert the TX and RX latency to 16 bits nanosecond and 16 bits fractional nanosecond format by multiplying them by 216 or 65536.
  3. Calculate the sum of the latency values and the TX/RX PMA delay values (In nanoseconds and fractional nanoseconds).
  4. Program the calculated 16 bits values to Triple-Speed Ethernet register.
    1. Program lower 16 bits TX latency values to TSE MAC register 0xD1, which is the TX fns value.
    2. Program upper 16 bits TX latency values to TSE MAC register 0xD2, which is the TX ns value.
    3. Program lower 16 bits RX latency values to TSE MAC register 0xD4, which is the RX fns value.
    4. Program upper 16 bits RX latency values to TSE MAC register 0xD5, which is the RX ns value.
Level Two Title