F-tile Architecture and PMA and FEC Direct PHY IP User Guide

ID 683872
Date 10/15/2021
Public

A newer version of this document is available.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

Document Table of Contents

3.5. Bit Mapping for PMA and FEC Mode PHY TX and RX Datapath

The tx_parallel_data bit and rx_parallel_data bit width depends on the PMA width and Number of PMA lanes IP parameters. Use the following equation to determine the total tx_parallel_data or rx_parallel_data bit width:

25

Total tx_parallel_data or rx_parallel_data Bit Width Equation:

tx/rx_parallel_data[(80*N*X)-1:0]

Where:

  • N = Number of PMA lanes value from 1 to 16.
  • X = Number of streams for the PMA configuration. Depending on PMA width, X can be 1, 2, or 4.

Refer to Table 1 for full variable definitions.

The tx/rx_parallel_data signals include the valid parallel data bits and other functionality bits, such as the data valid bit, the write enable for TX core interface FIFO in elastic mode bit, the RX deskew bit, and the alignment market bits (for FEC mode). These signals travel to and from the FPGA fabric to the F-tile, and are clocked by the same parallel clock. This parallel clock can be a PMA clock or System PLL clock.

Example 1: Total tx/rx_parallel_data Bit Width with 2 PMA Lanes (N=2) and 8-bit PMA Width (X=1)

tx_parallel_data [(80*2*1)-1:0] = tx_parallel_data [159:0]
rx_parallel_data [(80*2*1)-1:0] = rx_parallel_data [159:0]

Example 2: Total tx/rx_parallel_data Bit Width with 4 PMA Lanes (N=4) and 64-bit PMA Width (X=2)

tx_parallel_data [(80*4*2)-1:0] = tx_parallel_data [639:0]
rx_parallel_data [(80*4*8)-1:0] = rx_parallel_data [639:0]

Parallel Data Mapping information for TX and RX

If the PMA width is less than or equal to 32, D=PMA width.

If the PMA width is 64 or 128, D=32.

The lower case x is defined as x=0 to X-1. For a given lane, n and given stream x, you can calculate the TX and RX parallel data information according to the following tables:

Table 51.  PMA Direct Mode TX Parallel Data Information Calculations (Enable Double width transfer = 1)
TX Parallel Data MSB LSB
Write Enable for TX Core FIFO in Elastic Mode 26 79 + (80 * x) +(80 *n * X)
TX Data (Upper Data bits) (40 + D-1) + (80 * x) + (80 *n * X) 40 + (80 * x) + (80 *n * X)
TX PMA Interface Data Valid Bit 27 28 38 + (80 * x) + (80 *n * X)
TX Data (Lower Data bits) D-1 + (80 * x) + (80 *n * X) 0 + (80 * x) +(80 *n * X)
Table 52.  PMA Direct Mode RX Parallel Data Information Calculations (Enable Double width transfer = 1)
RX Parallel Data MSB LSB
Data valid for RX Core FIFO in Elastic Mode (25) 79 + (80 * x) + (80 *n * X)
RX Deskew 29 78 + (80 * x) + (80 *n * X)
RX Data (Upper Data bits) (40 + D-1) + (80 * x) + (80 *n * X ) 40 + (80 * x) + (80 *n * X)
RX PMA Interface Data Valid Bit (26) 38 + (80 * x) + (80 *n * X)
RX Data (Lower Data bits) D-1 + (80 * x) + (80 *n * X) 0 + (80 * x) + (80 *n * X)
Table 53.  PMA Direct Mode TX Parallel Data Information Calculations (Enable Double width transfer = 0)
TX Parallel Data MSB LSB
Write Enable for TX Core FIFO in Elastic Mode (25) 79 + (80 *n)
TX PMA Interface Data Valid Bit (26) (27) 38 + (80 *n)
TX Data D-1 + (80 *n) 0 + (80 *n)
Table 54.  PMA Direct Mode RX Parallel Data Information Calculations (Enable Double width transfer = 0)
RX Parallel Data MSB LSB
Data valid for RX Core FIFO in Elastic Mode (25) 79 + (80 *n)
RX PMA Interface Data Valid Bit (26) 38 + (80 *n)
RX Data D-1 + (80 *n) 0 + (80 *n)
Table 55.  FEC Direct Mode TX Parallel Data Information Calculations (Enable Double width transfer = 1)
TX Parallel Data MSB LSB
Alignment Marker30 77 + (80 * x) +(80 *n * X)
TX Data (Upper 33 bits) 72 + (80 * x) + (80 *n * X) 40 + (80 * x) + (80 *n * X)
TX PMA Interface Data Valid Bit (26) (27) 38 + (80 * x) + (80 *n * X)
Alignment Marker (25) 37 + (80 * x) + (80 *n * X)
TX Data (Lower 31 bits) 32 + (80 * x) + (80 *n * X) 2 + (80 * x) + (80 *n * X)
Sync Head 1 + (80 * x) + (80 *n * X) 0 + (80 * x) + (80 *n * X)
Table 56.  FEC Direct Mode RX Parallel Data Information Calculations (Enable Double width transfer = 1)
RX Parallel Data MSB LSB
RX Deskew 31 78 + (80 * x) + (80 *n * X)
RX Data (Upper 33 bits) 72 + (80 * x) + (80 *n * X) 40 + (80 * x) + (80 *n * X)
RX PMA Interface Data Valid Bit (26) 32 38
Alignment Marker (29) 37
RX Data (Lower 31 bits) 32 + (80 * x) + (80 *n * X) 2 + (80 * x) + (80 *n * X)
Sync Head 1 + (80 * x) + (80 *n * X) 0 + (80 * x) + (80 *n * X)
25 This section explains the bit mapping of TX and RX parallel data if the Provide separate interface for each PMA option is disabled. If the Provide separate interface for each PMA option is enabled, refer to the introduction of Signal and Port Referencex_is to view the bit mapping differences.
26 Applicable only when using PMA clocking mode only and when TX/RX core FIFO is in elastic mode.
27 Applicable only when using System PLL clocking mode.
28 For all bonded configurations, all TX PMA Interface Data Valid bits must be asserted at the same cycle of tx_coreclkin clock.
29 Applicable only when using PAM4 and X=2 or 4
30 The two alignment markers in this table must be driven together by the same signal.
31 Applicable only when using NRZ/PAM4 when X=2 or 4 or N > 1
32 Only one per system