To minimize jitter when using the Low Latency 10G MAC IP core on Intel® Arria® 10 devices, it is important to ensure that the advanced transmit (ATX) phase-locked loop (PLL) and the fractional PLL (fPLL) be placed so that they can source the input reference clock directly from the reference clock buffer without passing through the reference clock network.
For best jitter performance, Intel recommends placing the reference clock as close as possible to the transmit PLL.
Use a dedicated reference clock pin in the same transceiver bank.
There are two dedicated reference clock (refclk) pins available in each transceiver bank. The bottom refclk pin directly feeds the bottom ATX PLL, fPLL, and CMU PLL. The top refclk pin directly feeds the top ATX PLL, fPLL, and CMU PLL.
Use a location constraint to ensure that the ATX PLL and fPLLs are located in the optimal top or bottom location, aligned with your chosen dedicated refclk pin location.