C++ High Performance

Performance evaluation

To evaluate the performance benefits, we'll utilize the following code, which compares 100'000'000 strings:

auto n = size_t{100'000'000}; 
auto a = std::vector<String>{}; 
auto b = std::vector<String>{}; 
auto c = std::vector<String>{};a.resize(n);b.resize(n);c.resize(n);
// {Fill the vectors with random strings...} 
auto num_equal = 0; 
for(size_t i = 0; i < n; ++i) { 
  num_equal += (a[i] + b[i] == c[i]) ? 1 : 0; 
} 

When executing this on an Intel i7 7700k CPU, we get the following results:

Comparison type	Time (milliseconds)	Speed up
No proxy object	675	1.00 x
With a proxy object	62	10.7 x

In other words, the speed up with a proxy object is almost eleven times as fast, when we got rid of ...