What makes Corundum special? Mainly that you can add client code into it and maybe free up an isolated core, or stage packets to send on short notice.
Mainly of interest for low-latency finance and for high-performance compute clusters.
What makes Corundum special? Mainly that you can add client code into it and maybe free up an isolated core, or stage packets to send on short notice.
Mainly of interest for low-latency finance and for high-performance compute clusters.
Supporting critical traffic with TSN is a two steps process. First, you synchronize all the participating network nodes. For this you can use PTP (IEEE 1588), which is like an Ethernet level NTP (grossly oversimplified, but you get the idea). Once all the nodes are in sync, they can use time aware scheduling (TAS) where a TDM frame is overlaid over all the LAN, with Ethernet traffic classes (TC) assigned to specific ranges. In other words, you define a repeating pattern, split into different sequential zones, and TC are aligned to some zones. The goal is to define repeating ranges dedicated to specific traffic classes, where one can control the load and make sure there is no contention and traffic will go through with deterministic latency.
All this could be used in a plant, to support both best effort traffic but also sensitive real-time traffic for automation, while protecting the later.
TSN started out for media applications (broadcasting) over Ethernet, but is getting into industrial applications (see https://opcfoundation.org/).
Support for TSN is planned for 5G (NR) release 16, to support industrial applications.
All this area is in flux, so having a flexible programmable platform can be interesting.
TDMA is basically a simple demonstration scheduler that enables and disables queues on microsecond timescales, based on PTP time. One of the original reasons for building Corundum was to enable optical switching research, where data transmission into the switch must be precisely coordinated with the configuration of the switch itself. We have tried to do this in software, but the precision is limited and the CPU overhead is high. With corundum, the schedule is enforced in hardware, so it is extremely precise and does not add any CPU overhead.
Emulating such services in a packet-switched network requires very precise timing hence the reference to it in terms of IEEE1588 / Precision Time Protocol.