Elephant Flows are small number of flows carrying majority of network traffic. These heavy flows are introduced in the network time to time. 4G network is in existence and being upgraded to 5G network, but the devices used as hops and nodes are rarely upgraded. This poses a challenge of processing the instantaneous load, required to service the flow, with existing devices. The article proposes a method Resource Aware Parallel Computing (RAPC) that performs high performance computing by taking the hardware resources into consideration while execution. The experiments show that the result is significant with the performance boost of 20% as compared to Generic Parallel Computing (GPC) method.

Flows, 5G, HPC, Data Stream Processing System, Resource Aware Computing, Parallel Computing, GPU, NVIDIA, CUDA