int main()
{
	vector<thread> threads;
	threads.reserve(8);
	mutex m;

	vector<int> ints;
	ints.reserve(8000000);
	for (int i = 0; i < 800000; ++i)
		ints.push_back(1);

	int sum = 0;
	auto start = chrono::high_resolution_clock::now();

	for (int i = 0; i < 800000; ++i)
		sum += ints[i];

	auto end = chrono::high_resolution_clock::now();
	auto diff = end - start;

	cout << "index loop: " << chrono::duration<double, milli> (diff).count() << "ms" << endl;

	sum = 0;
	start = chrono::high_resolution_clock::now();

	for (auto& val : ints)
		sum += val;

	end = chrono::high_resolution_clock::now();
	diff = end - start;

	cout << "range loop: " << chrono::duration<double, milli>(diff).count() << "ms" << endl;

	sum = 0;
	start = chrono::high_resolution_clock::now();

	for (auto it = ints.begin(); it != ints.end(); ++it)
		sum += *it;

	end = chrono::high_resolution_clock::now();
	diff = end - start;

	cout << "iterator loop: " << chrono::duration<double, milli>(diff).count() << "ms" << endl;
	sum = 0;
	vector<chrono::duration<double, milli>> times;
	times.resize(8);

	auto func = [&](int start, int total, int index)
	{
		int partial_sum = 0;

		auto s = chrono::high_resolution_clock::now();
		for (int i = start; i < start + total; ++i)
			partial_sum += ints[i];
		auto e = chrono::high_resolution_clock::now();
		auto d = e - s;

		m.lock();
		cout << "thread " + to_string(index) + ": " << chrono::duration<double, milli>(d).count() << "ms" << endl;
		sum += partial_sum;
		m.unlock();

		times[index] = chrono::duration<double, milli>(d);
	};

	for (int i = 0; i < 8; ++i)
		threads.push_back(thread(func, i * 100000, 100000, i));

	for (int i = 0; i < 8; ++i)
		threads[i].join();

	chrono::duration<double, milli> total;
	for (int i = 0; i < 8; ++i)
		total += times[i];

	cout << threads.size() << " threads total: " << total.count() << "ms" << endl;
	cout << sum << endl;
}