[DblPlusUngood]

Concerning your first question, that experiment is not intended to measure the cost of GC, which can be made small by trading memory. The point of this experiment is to measure the unavoidable costs of the language.

[Someone]

I would think that measures less costs of the language than costs caused by differences in compiler quality/focus.

Or are there go features that make it inherently slower than C for kernel programming, even when ignoring GC?