Description
In your benchmark results (https://aka.ms/aspnet/benchmarks), there are many machines listed like Intel Linux, AMD Linux, Azure Arm64 Linux, etc. Apart from the cpu, I'm assuming all of these machines have different parts like networking chipsets, drivers, motherboards.
Currently, I believe you are tracking performance within a single machine and not making comparisons between different machines. I'm curious what would be the best way to make cpu performance comparisons between different machines while running ASP.NET if they're using different parts, especially network chipsets? Ie, does server machine A perform better than server machine B on a given ASP.NET benchmark?
Would it be recommended to run the whole benchmark (client and server) on the same machine? That way, the effects of different networking parts could be eliminated, but this also introduces other factors such as resource contention between the client and server.
Thanks