[monocasa]

They're not that disparate; the workloads are normally very dependent on the low latency interconnect of most supercomputers.