Connecting to Transceivers Using Block Automation for Versal ACAPs - 8.11 English

CPRI LogiCORE IP Product Guide (PG056)

Document ID
PG056
Release Date
2022-11-16
Version
8.11 English

Block Automation is the methodology in IP integrator which is used to connect the CPRI core to a Versal ACAP Transceiver. This is the only supported flow to correctly generate and configure the transceiver. Block Automation is also required to enable transceiver sharing with other IP cores. This Figure shows the Run Block Automation link in the IP integrator window.

Figure 5-6: Block Automation Link for Versal ACAPs

X-Ref Target - Figure 5-6

cpri-run-ba.PNG

This Figure shows the Block Automation dialog box. In this case, two CPRI IP cores have been placed onto the IP integrator canvas. The dialog box gives you the choice of selecting one or both cores for the application of Block Automation. There is also the option of Auto, Start_With_New_Quad, or Customized_Connections.

Figure 5-7: Block Automation Dialog Box

X-Ref Target - Figure 5-7

cpri-ba-dialog.PNG

If Auto is selected, the Block Automation will sequentially place the selected IP cores in the first available GT Quad locations that are suitable. If no existing Versal ACAP Transceiver GT Quad is available or suitable, then a new GT Quad will be created. Block Automation will try to pack CPRI cores (and other IP cores) into as few GT resources as possible. IP cores will share PLL resources when they have common settings (line rate, reference clock, and PLL type) or use different resources when they do not.

If Start_With_New_Quad is selected, the Block Automation will sequentially place the selected IP cores in new Versal ACAP Transceiver GT Quad IPs.

If Customized_Connections is selected, the Block Automation will allow the user to choose the specific GT locations to use for each IP core. This option will not add new GT Quads to the IPI canvas, so requires the GT Quads to be added to the IPI canvas prior to running Block Automation. This option gives the maximum flexibility around the wiring of IP cores to GT Quad locations.