Ethernet Design Example Components User Guide

RSS

UG-20019 | 2019.07.01

Time-of-day Clock

The Time-of-day (TOD) Clock streams 96-bit and 64-bit time-of-day to one or more timestamping units in an IEEE 1588v2 solution. The time-of-day consist of the following fields.

Field	96-bit Timestamp Format	64-bit Timestamp Format
Second	48 bits	—
Nanosecond	32 bits	48 bits
Fractional nanosecond	16 bits	16 bits

This component supports coarse and fine adjustments, and period correction. It also supports configurable period adjustment and offset adjustment.

You can instantiate the TOD clock through the Ethernet IEEE 1588 Time of Day Clock Intel^® FPGA IP in the Intel^® Quartus^® Prime software.

Release Information

IP versions are the same as the Intel^® Quartus^® Prime Design Suite software versions up to v19.1. From Intel^® Quartus^® Prime Design Suite software version 19.2 or later, IP cores have a new IP versioning scheme.

The IP versioning scheme (X.Y.Z) number changes from one software version to another. A change in:

X indicates a major revision of the IP. If you update your Intel Quartus Prime software, you must regenerate the IP.
Y indicates the IP includes new features. Regenerate your IP to include these new features.
Z indicates the IP includes minor changes. Regenerate your IP to include these changes.

Table 1. Ethernet IEEE 1588 Time of Day Clock Intel^® FPGA IP Release Information
Item	Description
Version	19.2.0
Release Date	July 2019
Supported Devices	Arria^® V GX/GT/GZ/SX/ST Intel^® Arria^® 10 GX/GT/SX Cyclone^® V SE/SX/ST Intel^® MAX^® 10 Stratix^® V GX/GT Intel^® Stratix^® 10 Intel^® Cyclone^® 10 GX

Resource Utilization

Table 2. Estimated Resource Utilization in Stratix^® V Devices
Configuration	ALMs	Combinational ALUTs	Logic Registers
PERIOD_CLOCK_FREQUENCY = 0 OFFSET_JITTER_WANDER_EN = 0	727	733	1041
PERIOD_CLOCK_FREQUENCY = 1 OFFSET_JITTER_WANDER_EN = 0	764	736	1224
PERIOD_CLOCK_FREQUENCY = 0 OFFSET_JITTER_WANDER_EN = 1	1782	2406	2297

Table 3. Estimated Resource Utilization in Intel^® Arria^® 10 Devices (10AX115U2F45I2SGES)
Configuration	ALMs	Combinational ALUTs	Logic Registers
PERIOD_CLOCK_FREQUENCY = 0 OFFSET_JITTER_WANDER_EN = 0	716	741	942
PERIOD_CLOCK_FREQUENCY = 1 OFFSET_JITTER_WANDER_EN = 0	689	734	1233
PERIOD_CLOCK_FREQUENCY = 0 OFFSET_JITTER_WANDER_EN = 1	1961	2735	2530

Configuring the TOD Clock

In the Intel^® Quartus^® Prime software, instantiate the TOD clock by selecting Ethernet IEEE 1588 Time of Day Clock Intel^® FPGA IP from the IP Catalog or Platform Designer (Interface Protocols > Ethernet > Reference Design Components). Specify the following parameters.

Table 4. TOD Clock Parameters Description
Name	Value	Default Value	Description
`PERIOD_CLOCK_FREQUENCY`	0 or 1	1	Set this parameter to 0 if the MAC connected to the TOD clock requires low period clock frequency, such as the Triple-speed Ethernet or legacy 10G Ethernet MAC. For this setting, the nanosecond field in the `Period` and `AdjustPeriod` registers is 9 bits wide. Set this parameter to 1 if the MAC connected to the TOD clock requires high period clock frequency, such as Low-latency 10G Ethernet or 40G/100G Ethernet MAC. For this setting, the nanosecond field in the `Period` and `AdjustPeriod` registers is 4 bits wide.
`OFFSET_JITTER_WANDER_EN`	0 or 1	0	Set this parameter to 1 to enable the offset, jitter, and wander timers; 0 otherwise. This parameter is enabled only when `PERIOD_CLOCK_FREQUENCY` is set to 0.
`DEFAULT_NSEC_PERIOD`	0 – n	0x0006	The reset value of the nanosecond field in the `Period` register. n is 0xF if the nanosecond field is 4 bits wide. Otherwise, n is 0x1FF.
`DEFAULT_FNSEC_PERIOD`	0 – 0xFFFF	0x6666	The reset value of the fractional nanosecond field in the `Period` register.
`DEFAULT_NSEC_ADJPERIOD`	0 – n	0x0006	The reset value of the nanosecond field in the `AdjustPeriod` register. n is 0xF if the nanosecond field is 4 bits wide. Otherwise, n is 0x1FF.
`DEFAULT_FNSEC_ADJPERIOD`	0 – 0xFFFF	0x6666	The reset value of the fractional nanosecond field in the `AdjustPeriod` register.

Using the TOD Clock

Follow these guidelines when using the TOD clock:

96-bit timestamp format—load the time-of-day using the time_of_day_96b_load_data[] bus or the SecondsH, SecondsL, and NanoSec registers. The bus value always takes precedence over the register values. When loading the time-of-day through the time_of_day_96b_load_data[] bus, the output is available in the time_of_day_96[] bus after one clock cycle. Hence, Intel^® recommends that you add one clock cycle to the value of the time_of_day_96b_load_data[] bus to accommodate the latency.
64-bit timestamp format—load the time-of-day using the time_of_day_64b_load_data[] bus. The output is available in the time_of_day_64[] bus after one clock cycle. Hence, Intel^® recommends that you add one clock cycle to the value of the incoming time-of-day to accommodate the latency.
The TOD clock does not synchronize the 96-bit and 64-bit timestamp format.
The drift, jitter, and wander timers restart each time a new time-of-day is loaded, either through the signal or configuration registers.

Adjusting TOD Drift

You can use the DriftAdjust and DriftAdjustRate registers to correct drifts in the TOD clock due to insufficient binary representation of the 16-bit fractional nanosecond field in the Period register.

For example, the Period register is set to 6.4ns for a 10G Ethernet application. The hexadecimal representation of this value is 0x6 ns and 0x6666.4 fns. The fractional nanosecond value, 0x0000.4, cannot be represented in 16 bits thus causing the time of day to drift from the actual time by 0x0002 fns every 5 clock cycles. In other words, the time of day drifts 953.6ns every 1 second. To correct this situation, configure the registers as follow:

DriftAdjust = 0x02, which sets the nanosecond field to 0x0 and the fractional nanosecond field to 0x2.
DriftAdjustRate = 0x5.

Adjusting Offset, Jitter, and Wander

The TOD clock supports several types of adjustments:

Offset—you can use the OffsetNS and OffsetFNS registers to adjust large offsets in assisting faster system convergence. The offset can be positive or negative. The maximum correction is (10⁹ - 50) ns.
Jitter—you can use the JitterTimer and JitterAdjust registers to achieve small time scales (milliseconds or microseconds) frequency correction.
The jitter adjustment can either be a positive or negative adjustment per unit time. This helps achieve better frequency corrections. For very low values of the jitter, such as 1 ns correction for every second, the timer must be larger and the adjustment value must be smaller.
For example, to achieve 1 ns correction every second in a clock domain of 3.2 ns period, configure the registers as follow:
- JitterTimer = 0x12A05F20, which is the hexadecimal value of (1000000000/3.2).
- JitterAdjust = 0x10000, which sets the nanosecond field to 0x1 and the fractional nanosecond field to 0x0.
Wander—you can use the WanderTimeLSB, WanderTimeMSB, and WanderAdjust registers to achieve large time scale correction.
The wander adjustment can either be a positive or negative adjustment per unit time. Wander adjustments are typically on larger time scales such as per hour. For very low values of the wander such as 1 ns per 24 hours, the timer must be larger and the adjustment value must be smaller.
For example, to achieve 1 ns correction every 24 hours in a clock domain of 3.2 ns period, configure the registers as follow:
- WanderTimerLSB[29:1] = 0x2D68_B000
- WanderTimerMSB[15:0] = 0x06239

Correcting TOD Offset

You can use the AdjustPeriod and AdjustCount registers to correct TOD offset. AdjustPeriod register value is used as effective period value when AdjustCount register is a non-zero value.

For example, to correct the TOD offset of 'a' ns by increasing the value of 'a' for a TOD clock running at 125 MHz (Period register value is 0x8 ns):

set AdjustPeriod to '8+b' ns
set AdjustCount to 'c'

where 'a = b * c' is fulfilled.

By setting a = 16 ns, b = 2 ns, and c = 8, the logic produces TOD with effective period value of 10 ns in the next 8 clock cycles. After 8 clock cycles, the logic resumes the normal operation where effective period value = 8 ns.

Similarly, to correct negative TOD offset, set AdjustPeriod to '<Period register value> – b'.

Interface Signals

Figure 1. Interface Signals of TOD Clock

Avalon-MM Signals

Table 5. Avalon-MM Signals Description
Name	Direction	Width	Description
`csr_address[]`	In	n	Use this bus to specify the register address you want to read from or write to. By default, the width of this signal is 4. When the `OFFSET_JITTER_WANDER_EN` parameter is set to 1, the width of this signal is 5.
`csr_read`	In	1	Assert this signal to request a read.
`csr_readdata[]`	Out	32	Data read from the specified register.
`csr_write`	In	1	Assert this signal to request a write.
`csr_writedata[]`	In	32	Data to be written to the specified register.
`clk`	In	1	Clock for the Avalon-MM interface, whose frequency is not more than 100 MHz.
`rst_n`	In	1	Active-low reset signal for the `clk` domain. Synchronous to `clk`.

Time-of-day Signals

Table 6. Time-of-day Signals Description
Name	Direction	Width	Description
`time_of_day_96[]`	Out	96	96-bit time of day streamed by the TOD clock.
`time_of_day_64[]`	Out	64	64-bit time of day streamed by the TOD clock.
`time_of_day_96b_load_valid`	In	1	Assert this signal for one clock cycle to indicate that the `time_of_day_96b_load_data[]` bus is valid. It indicates that the 96-bit time of day is synchronized and loaded into the TOD clock.
`time_of_day_64b_load_valid`	In	1	Assert this signal for one clock cycle to indicate that the `time_of_day_64b_load_data[]` bus is valid. It indicates that the 64-bit time of day is synchronized and loaded into the TOD clock.
`time_of_day_96b_load_data[]`	In	96	96-bit time of day from the master TOD clock.
`time_of_day_64b_load_data[]`	In	64	64-bit time of day from the master TOD clock.
`period_clk`	In	1	Clock for the TOD clock. Ensure that this clock is in the same clock domain as the TX and RX clock signals of the MAC IP.
`period_rst_n`	In	1	Active-low reset signal for the `period_clk` domain. Synchronous to `period_clk`.

Configuration Registers

Table 7. Register Description
Byte Offset	Name	Description	Access	HW Reset Value
0x00	`SecondsH`	The upper 16-bit second field of the 96-bit TOD. The value occupies bits 0 to 15. Bits 16 to 31 are not used. Read the TOD registers in this sequence: `NanoSec`, `SecondsL`, and `SecondsH`. Write the TOD registers in this sequence: `SecondsH`, `SecondsL`, and `NanoSec`. Reading the `SecondsH`, `SecondsL`, and `NanoSec` registers does not necessarily return the last values written to these registers.	RW	0x0
0x01	`SecondsL`	The lower 32-bit second field of the 96-bit TOD. To read from or write to the TOD registers, refer to the guidelines provided in the `SecondsH` register description.	RW	0x0
0x02	`NanoSec`	The 32-bit nanosecond field of the 96-bit TOD. Loading this register with a value equal to or larger than a billion leads to an incorrect timestamp. To read from or write to TOD registers, refer to the guidelines provided in the `SecondsH` register description.	RW	0x0
0x03	Reserved	–	–	–
0x04	`Period`	The period for the frequency adjustment. Bits [24:16]: The nanosecond field if the `PERIOD_CLOCK_FREQUENCY` parameter is set to 0. Bits [19:16]: The nanosecond field if the `PERIOD_CLOCK_FREQUENCY` parameter is set to 1. Bits [15:0]: The fractional nanosecond field. The remaining bits are not used. The reset value of this register, n, is determined by the value of the `DEFAULT_NSEC_PERIOD` and `DEFAULT_FNSEC_PERIOD` parameters.	RW	n
0x05	`AdjustPeriod`	The period for the offset adjustment. Bits [24:16]: The nanosecond field if the `PERIOD_CLOCK_FREQUENCY` parameter is set to 0. Bits [19:16]: The nanosecond field if the `PERIOD_CLOCK_FREQUENCY` parameter is set to 1. Bits [15:0]: The fractional nanosecond field. The remaining bits are not used. The reset value of this register, n, is determined by the value of the `DEFAULT_NSEC_PERIOD` and `DEFAULT_FNSEC_PERIOD` parameters.	RW	n
0x06	`AdjustCount`	Bits [31:20]: Not used. Bits [19:0]: The number of clock cycles used during offset adjustment.	RW	0x0
0x07	`DriftAdjust`	The value that the TOD clock uses to periodically adjust the time of day. Bits [31:20]: Not used. Bits [19:16]: The nanosecond field. Bits [15:0]: The fractional nanosecond field.	RW	0x0
0x08	`DriftAdjustRate`	Bit 31: The drift direction: 0 for addition and 1 for subtraction. Bits [30:16]: Not used. Bits [15:0]: The interval between drift adjustments in number of clock cycles. Writing a value other than 0 to this register triggers the drift adjustment.	RW	0x0
0x09	`OffsetNS`	Bit 31: Not used. Bit 30: The offset direction: 0 for addition and 1 for subtraction. Bits [29:0]: The nanosecond field of the offset. Writing a value other than 0 to this register triggers the offset in the time of day.	RW	0x0
0x0A	`OffsetFNS`	Bits [31:16]: Not used. Bits [15:0]: The fractional nanosecond field of the offset.	RW	0x0
0x0C	`JitterTimer`	Bit 31: Unused Bit 30: The direction of the jitter adjustment: 0 for addition and 1 for subtraction. Bits [29:0]: The timer value in number of clock cycles. Periodic jitter adjustment is disabled when this register is set to 0. Writing a value other than 0 to this register enables period jitter adjustment. Hence, write to this register last.	RW	0x0
0x0D	`JitterAdjust`	Bits [31:16]: The nanosecond field of the jitter adjustment. Bits [15:0]: The fractional nanosecond field of the jitter adjustment.	RW	0x0
0x10	`WanderTimerLSB`	Bit 31: Unused Bit 30: The direction of the timer adjustment: 0 for addition and 1 for subtraction. Bits [29:0]: The least significant byte of the timer in number of clock cycles. Writing a value other than 0 to this register enables wander timer adjustment. Hence, write to the `WanderTimerLSB` and `WanderTimerMSB` registers last.	RW	0x0
0x11	`WanderTimerMSB`	Bits [31:16]: Unused. Bits [15:0]: The most significant byte of the timer in number of clock cycles.	RW	0x0
0x12	`WanderAdjust`	Bits [31:16]: The nanosecond field of the wander adjustment. Bits [15:0]: The fractional nanosecond field of the wander adjustment.	RW	0x0

Time-of-day Synchronizer

The Time-of-day (TOD) Synchronizer provides an accurate synchronization between the time of day of a master TOD clock and a slave TOD clock. This component can synchronize the following combination of master and slave TOD clocks:

Master and slave TOD clocks that operate at the same frequency, between 125 MHz and 390.625 MHz. The synchronizer also supports different clock phases and PPM.
Master and slave TOD clocks that operate at different frequencies: 62.5 MHz, 125 MHz, 156.25 MHz, 312.5 MHz, or 390.625 MHz.

You can instantiate the TOD synchronizer through the Ethernet IEEE 1588 TOD Synchronizer Intel^® FPGA IP in the Intel^® Quartus^® Prime software.

Release Information

The IP versioning scheme (X.Y.Z) number changes from one software version to another. A change in:

X indicates a major revision of the IP. If you update your Intel Quartus Prime software, you must regenerate the IP.
Y indicates the IP includes new features. Regenerate your IP to include these new features.
Z indicates the IP includes minor changes. Regenerate your IP to include these changes.

Table 8. Ethernet IEEE 1588 TOD Synchronizer Intel^® FPGA IP Release Information
Item	Description
Version	19.2.0
Release Date	July 2019
Supported Devices	Arria^® V GX/GT/GZ/SX/ST Intel^® Arria^® 10 GX/GT/SX Cyclone^® V SE/SX/ST Intel^® MAX^® 10 Stratix^® V GX/GT Intel^® Stratix^® 10 Intel^® Cyclone^® 10 GX

Resource Utilization

Table 9. Estimated Resource Utilization in Intel^® Arria^® 10 Devices (10AX115U2F45I2SGES)
Configuration	ALMs	Combinational ALUTs	Logic Registers	Memory (M20K Blocks)
Default Mode	563	636	1678	3

Configuring the TOD Synchronizer

In the Intel^® Quartus^® Prime software, instantiate the TOD Synchronizer by selecting Ethernet IEEE 1588 TOD Synchronizer Intel^® FPGA IP from the IP Catalog or Platform Designer (Interface Protocols > Ethernet > Reference Design Components). Specify the following parameters.

Table 10. TOD Synchronizer Parameters Description
Name	Valid Values	Description
`TOD_MODE`	0, 1	Specifies the format of the time of day. 0: 64 bits. 48 bits nanosecond and 16 bits fractional nanosecond. 1: 96 bits. 48 bits seconds, 32 bits nanosecond and 16 bits fractional nanosecond. The default value is 1.
`SYNC_MODE`	0 – 15	Specifies the synchronization type between the master and slave TOD clocks. 0: 125-MHz master TOD clock and 156.25 MHz slave TOD clock. 1: 156.25-MHz master TOD clock and 125-MHz slave TOD clock. 2: The frequencies of the master and slave TOD clocks are the same, between 125 MHz and 390.625 MHz. This synchronization type supports different phase or PPM. Ensure that you also specify the period of the master and slave clocks using the `PERIOD_NSEC` and `PERIOD_FNSEC` parameters. 3: 156.25-MHz master TOD clock and 312.5-MHz slave TOD clock. 4: 312.5-MHz master TOD clock and 156.25-MHz slave TOD clock. 5: 125-MHz master TOD clock and 312.5-MHz slave TOD clock. 6: 312.5-MHz master TOD clock and 125-MHz slave TOD clock. 7: 125-MHz master TOD clock and 390.625-MHz slave TOD clock. 8: 390.625-MHz master TOD clock and 125-MHz slave TOD clock. 9: 156.25-MHz master TOD clock and 390.625-MHz slave TOD clock. 10: 390.625-MHz master TOD clock and 156.25-MHz slave TOD clock. 11: 312.5-MHz master TOD clock and 390.625-MHz slave TOD clock. 12: 390.625-MHz master TOD clock and 312.5-MHz slave TOD clock. 13: 125-MHz master TOD clock and 62.5-MHz slave TOD clock. 14: 156.25-MHz master TOD clock and 62.5-MHz slave TOD clock. 15: 312.5-MHz master TOD clock and 62.5-MHz slave TOD clock. The default value is 1.
`PERIOD_NSEC`	0 – 4'hF	Specifies the respective 4-bit nanosecond field for the reset value for the following clock frequencies: 125 MHz: Set this parameter to 4'h8 for 8 ns. 156.25 MHz: Set this parameter to 4'h6 for 6.4 ns. This value is the default value. 312.5 MHz: Set this parameter to 4'h3 for 3.2 ns. 390.625 MHz: Set this parameter to 4'h2 for 2.56 ns
`PERIOD_FNSEC`	0 – 16h'FFFF	Specifies the respective 16-bit fractional nanosecond field for the reset value for the following clock frequencies: 125 MHz: Set this parameter to 16'h0 for 8 ns. 156.25 MHz: Set this parameter to 16'h6666 for 6.4 ns. This value is the default value. 312.5 MHz: Set this parameter to 16'h3333 for 3.2 ns. 390.625 MHz: Set this parameter to 16'h8F5C for 2.56 ns
`SAMPLE_SIZE`	64, 128, or 256	Specifies the number of samples to use in calculating the FIFO buffer’s fill level. More samples results in a more accurate estimation of the fill level. However, the calculation time increases with the number of samples. The default value is 64.

Using the TOD Synchronizer

Figure 2. TOD Synchronizer in a Design

The TOD synchronizer uses a dual-clock FIFO buffer to receive the time of day from the master TOD clock and transmits it to the slave TOD clock. To ensure that the synchronization is accurate, the transfer latency must be taken into consideration. The sampling clock (clk_sampling) samples the fill level of the FIFO buffer and calculates the latency. Derive this clock signal from the same source as the master TOD clock or the slave TOD clock using a PLL.

Sampling Clock Frequency

To achieve the recommended frequency for the sampling clock, follow these steps:

The SYNC_MODE and SAMPLE_SIZE parameters determine the sampling clock factor, which is then used to determine the required PLL settings. Use the Table 11 to identify the sampling clock factor for your configuration.
Use the sampling clock factor identified in the previous step to determine the PLL settings. Table 12 lists the settings for Stratix^® V PLL Intel^® FPGA IP.

Table 11. Sampling Clock Factor
`SYNC_MODE`	Reference Clock Frequency (MHz)	Sampling Clock Factor
`SYNC_MODE`	Reference Clock Frequency (MHz)	`SAMPLE_SIZE` = 64	`SAMPLE_SIZE` = 128	`SAMPLE_SIZE` = 256
0, 1	125	16/63	32/33	64/63
0, 1	156.25	64/315	128/155	256/375
2	Master/slave frequency	64/63	128/153	256/375
3, 4	156.25	64/63	128/153	256/375
3, 4	312.5	32/33	64/63	128/153
5, 6	125	32/63	64/63	128/63
5, 6	312.5	64/315	128/155	256/375
7, 8	125	—	—	32/13
7, 8	390.625	—	—	256/375
9, 10	156.25	32/33	64/31	128/63
9, 10	390.625	64/155	128/185	256/375
11, 12	312.5	16/15	32/33	64/63
11, 12	390.625	64/75	128/185	256/253
13	62.5	64/63	128/153	256/253
13	125	32/33	64/63	128/153
14	62.5	32/33	64/63	128/153
14	156.25	64/155	128/155	256/375
15	62.5	64/63	128/153	256/253
15	312.5	64/155	128/155	256/375

Table 12. PLL Settings for Stratix^® V Devices
Sampling Clock Factor	PLL Counter
Sampling Clock Factor	M	N	C
16/15	16	5	3
16/63	16	3	21
32/13	32	13	1
32/33	32	3	11
32/33	32	11	3
32/63	32	3	21
64/31	64	31	1
64/63	64	9	7
64/63	64	21	3
64/75	64	25	3
64/155	64	31	5
64/315	64	21	15
128/63	128	21	3
128/153	128	51	3
	128	17	9
	128	9	17
128/155	128	31	5
128/185	128	37	5
256/253	256	11	23
256/375	256	75	5
	256	25	15

Interface Signals

Figure 3. Interface Signals of TOD Synchronizer

Table 13. Signals Description
Name	Direction	Width	Description
Clock and Reset Signals
`clk_master`	In	1	Master TOD clock domain.
`reset_master`	In	1	Synchronous reset signal in the master TOD clock domain.
`clk_slave`	In	1	Slave TOD clock domain.
`reset_slave`	In	1	Synchronous reset signal in the slave TOD clock domain.
`clk_sampling`	In	1	Sampling clock to measure the transfer latency.
Interface Signals
`start_tod_sync`	In	1	Assert this signal to start the synchronization process. Synchronization continues as long as this signal is asserted.
`tod_master_data[]`	In	64 or 96	Carries the 64-bit or 96-bit time of day from the master TOD clock. The width of this signal is determined by the `TOD_MODE` parameter
`tod_slave_valid`	Out	1	When asserted, the signal indicates that the data on the `tod_data_slave` bus is valid and ready for transfer in the following cycle. This signal stays asserted for only 1 clock cycle.
`tod_slave_data[]`	Out	64 or 96	Carries the 64-bit or 96-bit time of day for the slave TOD clock. This time of day is synchronized to the master TOD clock with an additional one clock cycle because it takes one clock cycle to transfer the time of day to the slave TOD clock. The width of this signal is determined by the `TOD_MODE` parameter.

Packet Classifier

The Packet Classifier decodes the packet type of incoming PTP packets, and returns the decoded information to the MAC IP core. The decoded information is aligned to the start of packet of the corresponding PTP packet.

You can instantiate the packet classifier through the Ethernet Packet Classifier Intel^® FPGA IP in the Intel^® Quartus^® Prime software.

Release Information

The IP versioning scheme (X.Y.Z) number changes from one software version to another. A change in:

X indicates a major revision of the IP. If you update your Intel Quartus Prime software, you must regenerate the IP.
Y indicates the IP includes new features. Regenerate your IP to include these new features.
Z indicates the IP includes minor changes. Regenerate your IP to include these changes.

Table 14. Ethernet Packet Classifier Intel^® FPGA IP Release Information
Item	Description
Version	19.2.0
Release Date	July 2019
Supported Devices	Arria^® V GX/GT/GZ/SX/ST Intel^® Arria^® 10 GX/GT/SX Cyclone^® V SE/SX/ST Intel^® MAX^® 10 Stratix^® V GX/GT Intel^® Stratix^® 10

Resource Utilization

Table 15. Estimated Resource Utilization in Intel^® Arria^® 10 Devices (10AX115U2F45I2SGES)
Configuration	ALMs	Combinational ALUTs	Logic Registers	Memory (M20K Blocks)
Default Mode	212	297	179	8

Configuring the Packet Classifier

In the Intel^® Quartus^® Prime software, instantiate the Packet Classifier by selecting Ethernet Packet Classifier Intel^® FPGA IP from the IP Catalog or Platform Designer (Interface Protocols > Ethernet > Reference Design Components). Specify the parameters in the following table.

Table 16. Packet Classifier Parameters Description
Name	Value	Default	Description
`TSTAMP_FP_WIDTH`	1 – 32	4	The width of the timestamp fingerprint.
`SYMBOLSPERBEAT`	1, 4, or 8	8	The number of symbols transferred in a clock cycle.
`BITSPERSYMBOL`	8	8	The number of bits per symbol transferred in a clock cycle.

Interface Signals

Clock and Reset Signals

Table 17. Clock and Reset Signals Description
Name	Direction	Width	Description
`clk`	In	1	Reference clock for the packet classifier. Connect this signal to the MAC TX clock.
`reset`	In	1	Synchronous reset signal for the packet classifier.

Avalon -ST Interface Signals

Table 18. Avalon^® -ST Signals Description
Name	Direction	Width	Description
`data_sink_sop`	In	1	Assert this signal to indicate the beginning of the packet.
`data_sink_eop`	In	1	Assert this signal to indicate the end of the packet.
`data_sink_valid`	In	1	Assert this signal to indicate that the `data_sink_data[]` signal and other signals on this interface are valid.
`data_sink_ready`	Out	1	When asserted, this signal indicates that the packet classifier is ready to accept data.
`data_sink_data[]`	In	n ¹	The input packet.
`data_sink_empty[]`	In	2, 3	Use this signal to specify the number of empty bytes in the cycle that contain the end of packet. The width of this signal is 2 when the SYMBOLSPERBEAT parameter is 4; 3 when the parameter is 8. This signal does not exist when the SYMBOLSPERBEAT is 1.
`data_sink_error`	In	1	Assert this signal to indicate that the current input packet contains errors.
`data_src_sop`	Out	1	When asserted, this signal indicates the beginning of the packet.
`data_src_eop`	Out	1	When asserted, this signal indicates the end of the packet.
`data_src_valid`	Out	1	When asserted, this signal indicates that the `data_src_data[]` signal and other signals on this interface are valid.
`data_src_ready`	In	1	Assert this signal when the receiving component is ready to accept data.
`data_src_data[]`	Out	n ¹	The output data.
`data_src_empty[]`	Out	2, 3	Contains the number of empty bytes in the cycle that contain the end of packet. The width of this signal is 2 when the SYMBOLSPERBEAT parameter is 4; 3 when the parameter is 8. This signal does not exist when the SYMBOLSPERBEAT is 1.
`data_src_error`	Out	1	When asserted, this signal indicates that the current output packet contains errors.

¹ n=SYMBOLSPERBEAT*BITSPERSYMBOL

Control Signals

Table 19. Control Signals Description
Name	Direction	Width	Description
`tx_etstamp_ins_ctrl_in_ingress_timestamp_96b`	In	96	The 96-bit timestamp received from the MAC RX, aligned to the start of packet of the corresponding PTP packet.
`tx_etstamp_ins_ctrl_in_ingress_timestamp_64b`	In	64	The 64-bit timestamp received from the MAC RX, aligned to the start of packet of the corresponding PTP packet.
`tx_etstamp_ins_ctrl_out_ingress_timestamp_96b`	Out	96	The 96-bit timestamp to the MAC TX, aligned to the corresponding output PTP packet.
`tx_etstamp_ins_ctrl_out_ingress_timestamp_64b`	Out	64	The 64-bit timestamp to the MAC TX, aligned to the start of packet of the corresponding output PTP packet.
`tx_egress_timestamp_request_in_valid`	In	1	Assert this signal to indicate that a timestamp is required for the packet. This signal must align to the start of an input packet.
`tx_egress_timestamp_request_in_fingerprint`	In	`TSTAMP_FP_WIDTH`	The timestamp's fingerprint for the input packet.
`tx_egress_timestamp_request_out_valid`	Out	1	Assert this signal when timestamp is required for the particular frame. This signal is aligned to the start of packet of the corresponding output PTP packet.
`tx_egress_timestamp_request_out_fingerprint`	Out	`TSTAMP_FP_WIDTH`	The timestamp's fingerprint for the output packet.
`clock mode`	In	2	Specify the clock mode: 00: Ordinary clock 01: Boundary clock 10: End to end transparent clock 11: Peer to peer transparent clock
`pkt_with_crc`	In	1	Use this signal to indicate whether or not the incoming packet contains 4-byte CRC field. 0: the incoming packet contains the CRC field. 1: the incoming packet does not contain the CRC field.
`tx_etstamp_ins_ctrl_in_residence_time_calc_format`	In	1	Use the following values to specify the format of the timestamp to use when calculating the residence time. 0: 96-bit timestamp 1: 64-bit timestamp Align this signal to the start of the input packet.
`tx_etstamp_ins_ctrl_out_residence_time_calc_format`	Out	1	The format of the timestamp used to calculate the residence time. 0: 96-bit timestamp 1: 64-bit timestamp This signal is aligned to the start of the output packet.
`tx_etstamp_ins_ctrl_out_checksum_zero`	Out	1	When asserted, indicates that the checksum field of the PTP packet is set to zero. This signal is aligned to the start of packet of the corresponding PTP packet.
`tx_etstamp_ins_ctrl_out_checksum_correct`	Out	1	When asserted, indicates that the checksum of the PTP packet is corrected by updating the checksum correction offset. This signal is aligned to the start of packet of the corresponding PTP packet.
`tx_etstamp_ins_ctrl_out_timestamp_format`	Out	1	The format of the timestamp. 0: 1588v2 format. 96-bit timestamp that consists of 48-bit second field, 32-bit nanosecond field, and 16-bit correction field for fractional nanosecond. 1: 1588v1 format. 64-bit timestamp that consists of 32-bit second field and 32-bit nanosecond field. This signal is aligned to the start of packet of the corresponding output PTP packet.
`tx_etstamp_ins_ctrl_out_timestamp_insert`	Out	1	When asserted, indicates that an egress timestamp must be inserted into the corresponding PTP packet. This signal is aligned to the start of packet of the corresponding PTP packet.
`tx_etstamp_ins_ctrl_out_residence_time_update`	Out	1	When asserted, indicates that the residence time is added to the correction field of the PTP packet. This signal is aligned to the start of packet of the corresponding PTP packet.
`tx_etstamp_ins_ctrl_out_offset_timestamp`	Out	16	The location of the timestamp field, relative to the first byte of the packet.
`tx_etstamp_ins_ctrl_out_offset_correction_field`	Out	16	The location of the correction field, relative to the first byte of the packet.
`tx_etstamp_ins_ctrl_out_offset_checksum_field`	Out	16	The location of the checksum field, relative to the first byte of the packet.
`tx_etstamp_ins_ctrl_out_offset_checksum_correction`	Out	16	The location of the checksum correction field, relative to the first byte of the packet.

Ethernet Design Example Components User Guide Archives

If an IP core version is not listed, the user guide for the previous IP core version applies.

IP Core Version	User Guide
18.0	Ethernet Design Example Components User Guide
16.0	Ethernet Design Example Components User Guide

Document Revision History for the Ethernet Design Example Components User Guide

Document Version	Intel^® Quartus^® Prime Version	IP Version	Changes
2019.07.01	19.2	19.2.0	Renamed the Supported Devices sections to Release Information in the respective chapters to include IP release information as well as supported devices. Added support for Intel^® Stratix^® 10 and Intel^® Cyclone^® 10 GX devices in the Time-of-day Clock and Time-of-day Synchronizer chapters. Added support for Intel^® Stratix^® 10 devices in the Packet Classifier chapter.
2018.12.03	18.0	–	Rebranded as Intel. Renamed the following Ethernet design example component names as per Intel rebranding: "Ethernet IEEE 1588 TOD Synchronizer" to "Ethernet IEEE 1588 TOD Synchronizer Intel FPGA IP" "Ethernet IEEE 1588 Time of Day Clock" to "Ethernet IEEE 1588 Time of Day Clock Intel FPGA IP" "Ethernet Packet Classifier" to "Ethernet Packet Classifier Intel FPGA IP" Updated the following topics: Time-of-day Clock Using the TOD Clock Updated Table: Register Description to correct the byte offsets of `SecondsL`, `NanoSec`, Reserved, `Period`, `AdjustPeriod`, `AdjustCount`, `DriftAdjust`, `DriftAdjustRate`, `OffsetNS`, `OffsetFNS`, `JitterTimer`, `JitterAdjust`, `WanderTimerLSB`, and `WanderAdjust` registers. Made editorial updates throughout the document.

Date	Version	Changes
March 2017	2017.03.08	Added a new topic: Correcting ToD Offset. Updated the "Time-of-day Signals Description" table. Added resource utilization for Arria 10 Devices for 16.1 release: "Time-of-day Clock" section "Time-of-day Synchronizer" section "Packet Classifier" section
May 2016	2016.05.02	Initial release.