For hosts with 40G/100G Ethernet NICs, there are some additional things you'll want to tune to maximize throughput.
The most important things to configure are:
- CPU governor to 'performance'
- TCP buffer size set to the maximum (2GB)
- make sure you are using the correct cores for both IRQ and user processes
- enable on 'fair queuing' (FQ) with a good pacing rate for your environment.
No other tuning should be needed for modern Linux OS's (RHEL/CentOS 7.2 and higher, or systems with a 4.x kernel).
For more details on 100G tuning, see this presentation from September 2016.
CPU clock rate still matters a lot for 40G/100G flows If you care about the throughput of single flows, a higher CPU clock rate is important. In general, you need a CPU clock rate of at least 3GHz to achieve 40Gbps per flow.
To find other publicly accessible 40/100G hosts to run tests with, see: