F-tile ethernet hard ip in agilex7
I generate 200G ethernet F-tile hard ip that base on agilex7, the IP configuration as follow: Our current test setup uses our custom development boards interconnected with Mellanox CX-8. The link fails to come up regardless of whether the CX-8 is configured in forced mode or auto-negotiation mode. Could you advise on how I should troubleshoot this issue next? Additionally, I would like to ask: Does the F-Tile Ethernet Hard IP support filtering packets smaller than 64 bytes? I cannot find the corresponding configuration options in the IP configuration GUI.F-tile ethernet hard ip in agilex7
I generate 200G ethernet F-tile hard ip that base on agilex7, the IP configuration as follow: Our current test setup uses our custom development boards interconnected with Mellanox CX-8. The link fails to come up regardless of whether the CX-8 is configured in forced mode or auto-negotiation mode. Could you advise on how I should troubleshoot this issue next? Additionally, I would like to ask: Does the F-Tile Ethernet Hard IP support filtering packets smaller than 64 bytes? I cannot find the corresponding configuration options in the IP configuration GUI.
Agilex 7 F-Tile 200G hard IP de-feature clarification
There is a point in the Agilex 7 known issues list in UG-683584 for some silicon revisions The F-Tile Ethernet 200G Hard IP block is de-featured and cannot beused in production devices with OPNs that have no suffix (blank) or "B" suffix Does this mean that 200G "bonded" hard-IP is not supported in those older OPNs, such as 200GE-4 / 200GE-8 ? Or does it mean that all parts of 200G hard IP is de-featured? For example would FEC/PCS/MAC still work in 10/25/50GE-1 configurations?Agilex7 F-TILE ethernet hard IP 200G
Now we develop a project with agilex7 fpga. In the project we generate 200Gx4 ethernet hard ip. during testing on-board, the reset flow of the ethernet IP failed to complete successfully: The IP output signals o_rst_ack_n / o_tx_rst_ack_n / o_rx_rst_ack_n remain permanently High and never pull Low. This causes the input reset signals i_rst_n / i_tx_rst_n / i_rx_rst_n fed into the Ethernet IP to stay stuck in the Low-level reset state and never exit reset. the following signal status as follow by read register: tx_lanes_stable = 0 sys_pll_locked = 0 tx_pll_locked = 1 the IP reset sequence in the IP datasheet shows: so what should we do for the next step? Do you have any suggestion?
25G Ethernet IP for Stratix10
Team, We have 25G Ethernet IP for stratix10 which suport 10G and 25G dynamically. How to set it to 10G rate and 25G as and when required. We understating it is done by writing to some registers through the reconfiguration interface. We cannot find these registers and the process to be followed to switch the 10G25G interface rates. Regards amolInterfacing Avalon Streaming FIFO IP with GTS Ethernet Hard IP (Agilex 5, quartus v25.3)
Hello, I posted this question before on the Quartus Prime Forum but saw this forum and thought maybe this would be a better place to post it. Sorry if this is considered spam and not allowed: I want to integrate the Avalon Streaming Single Clock FIFO IP (AVST FIFO IP) with the GTS Ethernet Hard IP (GTS EHIP) , but the GTS EHIP outputs signals rxstatus_valid and rxstatus_data, that don't interface with the AVST FIFO IP. The AVST FIFO IP is in a custom module that sits in between the GTS EHIP and the rest of the 1x10G Ethernet System Example Design: Agilex 5 FPGA E-Series Modular Development Kit (Link: https://altera-fpga.github.io/rel-25.3/embedded-designs/agilex-5/e-series/modular/ethernet/agx5e-ethernet-10g/ug-agx5e-ethernet-10g/). How should I handle these signals? Can I ignore them? Is there a example reference design that does this? Thank you for the help! IP Blocks (left:EHIP, right:avst sc fifo ip): SC FIFO Parameters:
F-Tile Ethernet Hard IP (100G)
We are using the "F-Tile Ethernet Hard IP" (rev 15.0.0) in 100GE-4 + FEC mode in an Agilex 7 since >1 year, and it has worked fine when we build it with Quartus 24.2, but now after upgrading to Quartus 25.3.1 (patch 1.02) our 100G Ethernet switch no longer gets a link to the FPGA. I have re-generated the"F-Tile Ethernet Hard IP" (rev 20.1.0) in Q25.3.1 with exactly the same settings (not using 'Advanced mode') as we had when generating it in Q24.2 One difference I see in the *.ip files is this: rev 15.0.0: CUSTOM_RATE_GUI 25.78125 rev 20.1.0: CUSTOM_RATE_GUI 10.3125 These settings are not visible when not running 'Advanced mode' Is there anyone else that has seen problems with this IP after upgrading to Q25.3.1? A side note, we are also using the IP in 400GE-4 mode, and that works fine both in Q24.2 and Q25.3.1Error when simulating F-tile Ethernet example design
When running "run_vcsmx.sh" following set of errors occurs Just one example Error-[URMI] Unresolved modules ../ex_200G/sim/../../hardware_test_design/support_logic/eth_f_hw_auto_tiles.sv, 2655 "intfc_m_hdpldadapt_avmm1_mux #(.topology("UX16E400GPTP_XX_DISABLED_XX_DISABLED"), .maib_id(0), .num_ip_on_intfc_00(2), .system_pll_ip_index_on_intfc_00(1), .num_ip_on_intfc_01(2), .system_pll_ip_index_on_intfc_01(1)) z1577b_x0_y0_n0__avmm1_0( .pld_avmm1_busy_real (z1577b_x0_y0_n0__avmm1_0__pld_avmm1_busy_real), .pld_avmm1_clk_rowclk_real (z1577b_x0_y0_n0__avmm1_0__pld_avmm1_clk_rowclk_real), .pld_avmm1_cmdfifo_wr_full_real (z1577b_x0_y0_n0__avmm1_0__pld_avmm1_cmdfifo_wr_full_real), .pld_avmm1_cmdfifo_wr_pfull_real (z1577b_x0_y0_n0__avmm1_0__pld_avmm1_cmdfifo_wr_pfull_real), .pld_avmm1_read_real (z1577b_x0_y0_n0__avmm1_0__pld_avmm1_read_real), .pld_avmm1_readdata_real (z1577b_x0_y0_n0__avmm1_0__pld_avmm1_readdata_real), .pld_avmm1_readdatavalid_rea ... " Module definition of above instance is not found in the design. eth_f_hw_auto_tiles.sv is a 36 MB sv-file, making debugging of this quite tedious. I've only followed the design example user guide which states the script should be able to run out of the box.
Agilex3/5 GTS Hard Ethernet IP 10G example design pin loc and io std wanted
Where can I find any public available dev kit design example for the Agilex3 or Agilex which can implement the GTS Eth HIP as generated by Quartus Pro v25.3 and successfully build a sof file? A set of pin locations and IO standard settings for the AXE5 Eagle would be optimal, but any other dev kit would be helpful. According to the "GTS Ethernet Hard IP User Guide: Agilex 3 FPGAs and SoCs" (848477) page 29 under "Target Development Kit Tab" is says: "Target development kit option specifies the target development kit used to generate the project. Ensure the pin assignments in the .qsf file are appropriate." But it seems like this will only set the BOARD parameter in the resulting qsf, e.g. when using the Premium Development Kit it results in the following addtion to the qsf file: set_global_assignment -name BOARD "Agilex 5 FPGA E-Series 065B Premium Development Kit DK-A5E065BB32AES1" which results in no location or IO standard settings in the qsf and I/O Assignment Warnings in the fitter report after the build: +-----------------------------------------------------------------------------------------------------------------------+ ; I/O Assignment Warnings ; +-----------------------+-----------------------------------------------------------------------------------------------+ ; Pin Name ; Reason ; +-----------------------+-----------------------------------------------------------------------------------------------+ ; o_tx_serial_data[0] ; Incomplete set of assignments. Missing I/O standard, drive strength and slew rate assignments ; ; o_tx_serial_data_n[0] ; Incomplete set of assignments. Missing I/O standard, drive strength and slew rate assignments ; ; qsfp_lowpwr ; Incomplete set of assignments. Missing I/O standard, drive strength and slew rate assignments ; ; qsfp_rstn ; Incomplete set of assignments. Missing I/O standard, drive strength and slew rate assignments ; ; i_reconfig_clk ; Missing I/O standard ; ; i_rx_serial_data_n[0] ; Missing I/O standard ; ; i_rx_serial_data[0] ; Missing I/O standard ; ; i_clk_ref_p ; Missing I/O standard ; ; o_tx_serial_data[0] ; Missing location assignment ; ; o_tx_serial_data_n[0] ; Missing location assignment ; ; qsfp_lowpwr ; Missing location assignment ; ; qsfp_rstn ; Missing location assignment ; ; i_reconfig_clk ; Missing location assignment ; ; i_rx_serial_data_n[0] ; Missing location assignment ; ; i_rx_serial_data[0] ; Missing location assignment ; ; i_clk_ref_p ; Missing location assignment ; ; i_refclk2pll_p ; Missing location assignment ; +-----------------------+-----------------------------------------------------------------------------------------------+ Whenever I try to assign these myself I get errors like Error (14566): The Fitter cannot place 1 periphery component(s) due to conflicts with existing constraints (1 IPFLUXTOP_UXTOP_WRAP(s)). Fix the errors described in the submessages, and then rerun the Fitter. The Intel FPGA Knowledge Database may also contain articles with information on how to resolve this periphery placement failure. Review the errors and then visit the Knowledge Database at https://www.intel.com/content/www/us/en/support/programmable/kdb-filter.html and search for this specific error message number. Error (175001): The Fitter cannot place 1 IPFLUXTOP_UXTOP_WRAP, which is within GTS Ethernet Hard IP ex_10G_intel_eth_gts_1000_6dyx4dq. or this or other type of layout or clocking type constraint errors: Error (11216): Output port "O_SYSPLL_C0" of "SM_HSSI_PLL_WRAP" cannot connect to PLD port "CLK" of "FF" for node "kr_dut|intel_eth_anlt_gts_0|ip_inst|sip_inst|u_intel_eth_anlt_gts_csr_top|u__intel_eth_anlt_gts_csr_avmm_arb|o_avmm_rdata[0]". It would be nice if I could obtain a set of correct and working pin assignment which actually results in a working sof file so I can try to understand what the actual constraints are. Is there a dev kit as described which the pin assignments are generated or provided or could anybody please provide a set of pin assignments for the above signals for a dev kit? Cheers!
