It may be assumed that the hardware contributors have included the best medium access control available. However, if this were the case there would be no need for middle hardware switch/routing because control could be distributed to the edge. The result will do for switching what Linux did for serving, in that the middle hardware should just become a commodity box.
To that end, we would be happy to contribute our high-performance MAC for the purpose of providing a migration path to the next-generation low level switch architecture that does Ethernet better than Ethernet. Distributed Queue Switch Architecture was originally invented for cable TV before DOCSIS was chosen as the standard. The goal was to simultaneous handle synchronous and asynchronous flows, hence the broadcast architecture which eliminates backoffs, ensuring stable QoS and strict determinism for media packets. Of course, energy per bit transferred and security are vastly improved. More recently, the DASH7 Alliance has demonstrated DQSA within ITU 18000-7, which is a next-gen IoT specification now performing at ~100% throughput. All we need to do now is increase the data rate for voice and video, but again, the result will be that the only limiting factor is physical bandwidth. Otherwise, DQSA is independent of the phy and requires little to no change at 2.5 or above. So does Open Compute have the ability to C code, compile, port, test and debug the existing com chip?
_______________________________________________ opencompute-networking mailing list Unsubscribe: http://lists.opencompute.org/mailman/options/opencompute-networking opencompute-networking@lists.opencompute.org http://lists.opencompute.org/mailman/listinfo/opencompute-networking
