While that has been true for 70 years, it is ceasing to be true now; today you can often make your program faster by causing it to do more computation in order to avoid communication, or in order to tolerate communication latency, which may involve doing more communication, or occasionally in order to tolerate node failures, so that you can take advantage of more nodes.
In the first case, you are doing less (communication) to go faster.
In the second case, you are doing less (computation) to go faster.
The less expensive operations you have to do (whatever they may be), the faster your program is. That has been true for the past 70 years, and will continue to be true as long as computing remains in the slightest bit recognizable to men of our field as it currently stands.
In the first case (to do less communication), you are sort of correct; but the amount of extra computation can be truly enormous, and the version of your program that does the least communication and computation is necessarily single-threaded — and so, in the case of a parallelizable problem, slower.
In the other two cases (doing extra work to tolerate communication latency and node failures) I have searched long and hard for a way that your statement could be correct, but I cannot imagine what it could be.
You are very likely correct that it continued to be true as long as computing remained recognizable to the men (and women) of earlier generations. But in a world of multiprocessor microcontrollers, it is no longer true.
Yes, but that definition you quote is for "man" (often written as the more archaic "mankind"), not "men". You're actually using definition a.1: "an individual human; especially : an adult male human", in the plural, to refer to adult male humans of our field. Admittedly a pedantic point, I wouldn't bother if it didn't contribute to a real-world problem.
Now take your self-righteous offtopicness and shoveoff. I contribute nothing to the problems of society by exercising freedom of vocabulary. You want to see real problems? Go look at pictures of the shelled remains of children in Syria. You're suggesting that I am contributing to a "real-world problem" is insulting.
Even if we want to consider the furthering of systemic sexism through benign use of vocabulary a "real-world problem", you are more to blame than I by trying to convert an innocent phrase into an example of hatred. Words only have the power that you allow them to. I accept no responsibility for your oversensitivity.
Meh, it's a smaller problem than most, but it's also far easier to solve than most. Why avoid making the world a tiny bit better if it costs you nothing? If making minor tweaks to our word choices could avoid violence to Syrian children, too, that'd be grand, but alas...
This is not exactly a response to you, BTW (since I can't imagine you'd be convinced) but for other readers.
I think that for the sake of backwards compatability with the linguistic upbringing of most English speakers, we should use "men" to refer to both human genders.
If this means that male gendered individuals will have to be pluralized as "males" and never "men", then that's something we should live with.
Does the claim make sense in newer processors which when halted switch to a low-power C-state? Waking them up can consume a lot of time (order of milliseconds).
Priceless...