Reducing frame latency thanks to Cut-through approach in our MES IP Core

In this new post, how Fast Frame Forwarding (Cut-through) works in our Ethernet switch, will be explained.

Our Ethernet Switch, MES IP Core, provides Fast Frame Forwarding also called Cut-through forwarding. This frame forwarding mechanism diff ers from store and forward approach. Cut-Through Mode allows for the reduction of frame latency through the switch by forwarding frames directly to the egress port without first waiting for receipt of the entire frame and jumping the standard switching path as it is depicted in the following figure:

MES allows each priority queue to be independently set to either cut-through or store-and-forward operation. In that sense, it is possible to manage some priority frames in store and forwarded mode, whereas cut-through mode is applied to other frames (usually the highest priority frames).


Cut-through approach, compared to store-and-forward approach, is lower latency. The amount of time the device takes to start forwarding the frame (referred to as the switch’s latency) is on the order of a few microseconds only, regardless of the frame size. Table 2 provides the latency details for Cut-through approach. For non-cut-through packets, the minimum latency is proportional to the size of the packet (see Table 1).

Table1.: Store and Forward latencies for diff erent frame lengths.

 

 Table2.: Cut-through latencies for different frame lengths.

This kind of approach is very interesting for Fieldbus use-cases like Profinet, Ethernet/IP, Modbus, etc. Daisy-Chain & Ring topologies would benefit from this reduced frame latency technique.


System-on-Chip engineering S.L. (SoCe) is a worldwide leading supplier of time-aware Ethernet networking solutions. SoC-e is pioneer in developing a portfolio of IP cores and rugged platforms that implement these technologies for critical systems.

For more information, please contact us at: info@soc-e.com