It is interesting about "mean" and "average".

I think that they are synonym here, "mean" is even closer to the point because it reflects how it works: for several cycles the running times are measured, then the measurements are summarized and divided by the number of cycles.

I do not understand this at all. You cannot have a "mean time" of **one **operation. A mean time is a function of the timings of numerous operations. (NB **A** mean time (countable) )

I understand you confusion, by "mean time" I mean the mean time of a set of measurements.

We run the machine cycle several times, write down the time of each execution and calculate the mean.

In "

*mean time of a machine cycle,*" "a machine cycle" is a generalization here. It can be rewritten as "

*This test measures the mean time of a set of machine cycles,"* but then there would be no questioning of the article, "the" would be the only choice.

Now I start to doubt whether this is a legitimate generalization, is it?

My concern about the "the" article in variant 1 is following:

We don't use "the" with indefinite objects, obviously.

"mean time of a machine cycle" should be read as "mean time of a set of measurements".

Each running of the test spawns its own measurements, so that the mean of the measurements can be different for different test executions. The mean varies with test executions, and therefore it should be indefinite in the general sense of a single execution. That's why I think variant 2 with "a mean" could be right.