Esote esote.net

Unsolved inequality

The following inequality resulted from Waring's problem. Paraphrased from Wikipedia:

Are there any positive integers k ≥ 6 such that:

3^{k}-2^{k}\left\lfloor \left({\tfrac
	{3}{2}}\right)^{k}\right\rfloor
	>2^{k}-\left\lfloor \left({\tfrac {3}{2}}\right)^{k}\right\rfloor
	-2

Mahler proved that there could only be a finite number of k; none are known.

Kubina and Wunderlich, in their book Extending Waring's conjecture to 471,600,000, have shown that any such k would need be larger than 471,600,000.

It is conjectured, but not proven, that no such k exist. I have been unable to find a name for this specific inequality. From what I have found, it is not called "Waring's problem" but instead is a result of Waring's problem.

Unsolved-inequality.cpp

Change the value of PRECISION to increase the maximum k (or more accurately the maximum 3^k that the program can reach. You can also change the starting k by changing the value of k = 6 below the beginning of the main function. This program is not optimized, and there are definitely better ways to brute-force this inequality.

This program was tested using the g++ compiler. I recommend using -Ofast. Below is a comparison of run times until precision is exceeded using the different optimization flags. While my system will be different from others, the relative time improvements are what is important.

Optimization Flag Real Time
None 11m 40.775s
-O1 (-O) 3m 59.017s
-O2 2m 47.888s
-O3 1m 50.419s
-Ofast 1m 47.860s

So, use -Ofast or -O3 unless you have time to kill.