is std::copy faster than std::cout for output?

Question

std::vector<int> v{2,4,6,8,10,12,14,16,18,20};

// print the numbers
    std::copy(v.cbegin(), v.cend(), std::ostream_iterator<int>(std::cout, " "));
    std::cout << '\n';

here std::copy is used to write to std out. Is this faster than using std::cout for the vector elements in a for loop like

for(auto element: v) std::cout << element << " ";

I could't find much information about how they would write to output buffers for both.

Did you try to profile it before ? I think you can easily find the answer by yourself :) — Fareanor, Oct 30 '19 at 08:38
I'm not sure about that. I need to know if they work differently; if so, how? — rohitt, Oct 30 '19 at 08:40
You could look at the created assembler code, if you want to know, how they work — RoQuOTriX, Oct 30 '19 at 08:42
This https://stackoverflow.com/questions/22387586/measuring-execution-time-of-a-function-in-c show how you can measure it — Support Ukraine, Oct 30 '19 at 08:42
Pretty sure the bottle neck will be the console itself anyway, so if there are differences, they most likely are neglectable... — Aconcagua, Oct 30 '19 at 08:49

Ton van den Heuvel · Answer 1 · 2019-10-30T10:05:30.803

To give an rough idea on the relative performance of the two, see the benchmark results here: http://quick-bench.com/wGYYPBXEgvLrkyp5gpJOnIpt7A4

I had to output to a std::stringstream instead of std::cout to keep quick-bench happy. It gives some insight on the raw performance of the underlying implementations, but not on how they crossplay with a highly OS dependent output stream like std::cout.

So it is hard to come to any definitive conclusions based on such a simple benchmark alone. I would take from this that in reality there is most likely not enough difference between the two approaches to prefer one over the other from a performance perspective.

is std::copy faster than std::cout for output?

1 Answers1

Linked