results.txt

Numbers indicate the amount of iterations in 5 seconds.

The C example was compiled using:

    clang -Wall -Wextra -pedantic -O2 test.c -o test
    
Results:

C:        2272688181
LuaJIT    170557787   => ~13x slower
Lua       86128887    => ~26x slower (no JIT)
JRuby 9k  25726121    => ~88x slower (Invoke Dynamic, no JIT threshold)
JRuby 9k  23399982    => ~97x slower
rbx git   18267317    => ~124x slower
MRI 2.2.3 10886124    => ~208x slower (no JIT)

test.c

#include <stdio.h>
#include <time.h>

int main() {
    volatile size_t a = 10;
    volatile size_t b = 5;

    size_t iterations = 0;

    int start = (int) time(NULL);

    while ( (((int) time(NULL)) - start) <= 5 ) {
        size_t res = (a + b + iterations);

        iterations++;
    }

    printf("Iterations: %lu\n", iterations);

    return 0;
}

test.lua

function measure()
    local a = 10
    local b = 5
    local start = os.time()
    local iterations = 0

    while (os.time() - start) <= 5 do
        local res = a + b + iterations

        iterations = iterations + 1
    end

    return iterations
end

measure()

print("Iterations: " .. measure())

test.rb

def measure
  a = 10
  b = 5
  start = Time.now
  iterations = 0

  while (Time.now - start) <= 5
    res = a + b + iterations

    iterations += 1
  end

  iterations
end

measure

puts "Iterations: #{measure}"