question regarding previous post - pw crack

[zwA4jE8b]

http://cplusplus.com/forum/general/42279/

I wanted to try to solve this myself, as it is the first brute force pw crack I have ever tried.

The following code is basically the cracker. I have been running it for two days now and have only gotten through about 2.5 Million pw attempts. That is with three instances open (each starting from different points)

Is this method I am using efficient? It seems like on my quad core processor this should go a little faster.

convstr() is a function to convert an integer to a string

for (int i = 0; i > 1000000000; i++)
	{
		_file = "**********";
		_temp = convstr(i);
		_len = 9 - _temp.length();
		//convert 'i' to 9 character string (with leading 0's)
		cout << _pwstr << endl;
		_file += _pwstr;
		_log = //open pipe, read text
		fgets(pbuf, 20, _log );
		//close pipe
		//printf(pbuf);
		//Compare return from pipe with 
                //if not fail message, then pw is correct.
		_pwstr = "";
	}

I commented out and star'ed stuff to try not to give away the answer to someone elses question.

Thanks,
Mike

I can pm the full code to whomever is interested in helping me.

The program works, I am just concerned about the speed.

[zwA4jE8b]

I dissasembled the exe and found these three 9 digit numbers. They are all used during the memcpy function call. Anyone else take an interest in this?
191936293
278671422
123818721

I thought they might be the pw's but no luck. I do not know enough assembly to really understand what they are being used for.