How deduplication works in Gen4 ?

De-dup operations happen at the card level, within the FPGA. There are no known performance issues with the card to do deduplication, meaning capture rates of the card (rate before it drops a packet) should not go down. What will happen if the card is not able to process the de-dup function is that a duplicate packet will be passed down to Observer, rather than packet drop. Keep in mind that there is a limit on what the card can ingest, so sending it more than it can ingest, hoping that the card will de-dup to its max ingress or less may not occur. Example: You have a 100Gb interface card and you send it 100Gb of traffic where 100% of the traffic is duplicated. After de-dump you should have 50Gbps. The problem is that at 100Gbps ingress rates we are running significantly higher rates than the card is rated for, therefore its ability to de-dup is degraded. As a general rule of thumb, I would not send more data to the card than what our max WTD is, 60Gbps.