Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

In my previous questions it is shown that $f(x)=x^2+ax+a$ , where $a\in\mathbf{Z^+}$\ $\left \{ 4 \right \}$ is irreducible and that gcd$(f(1),f(2),f(3).....)=1$

So, according to Bunyakovsky conjecture this polynomial produces infinitely many primes for all $a\in\mathbf{Z^+}$\ $\left \{ 4 \right \}$

My question : Can we find such coefficient $a$ that $x^2+ax+a$ is composite for all natural x ?

share|cite|improve this question
Does a = 4 count or am I missing something? – Sp3000 Nov 12 '11 at 6:21
@Sp3000,please read carefully question..$a$ is positive integer and $a \neq 4$ – pedja Nov 12 '11 at 6:25
Hm, I should learn my notation properly, sorry about that. Well unless my code's wrong I can't find any $a < 500000$ and almost all $a$ with a long chain of composites are even. – Sp3000 Nov 12 '11 at 6:36
@Sp3000,what programming language you use...? May you post your code as answer ? – pedja Nov 12 '11 at 6:40
I know no counterexample. As far as I know, no one does, or at least no one did a day or two ago. The news would spread quickly! – André Nicolas Nov 12 '11 at 7:02
up vote 1 down vote accepted

There is a good reason for the Bunyakovsky conjecture to be open, it is very hard to find a counterexample and even for polynomials that seem to be one, there are cases where you have to wait for very high $x$ until you encounter a prime. See for exampe the one given at wolfram: $x^{12}+488669$ it produces the first prime for $x=616980$, the first prime is therefore bigger than $10^{69}$.

Using GMP you can try if there might be a possible counterexample in your sequence:

#include <iostream>
#include <gmpxx.h>
using namespace std;

int Rabin_Miller(mpz_class n) { /* given an integer n >= 3 returns 0 composite, 1 probably prime */
    return mpz_probab_prime_p(n.get_mpz_t(), 10);

#define MAX_A 1000000
#define MAX_X 1000

int main() {
    mpz_class a;
    for(a = 4; a < MAX_A; ++a)
        for(unsigned int x = 1; x <= MAX_X; ++x)
            if(Rabin_Miller(x*x+a*x+a) > 0)
                x = MAX_X+1;    
            if(x == MAX_X)
                cout << "For a = " << a << " the first " << MAX_X << " values are composite!" << endl;

I used it to see (takes over 10 minutes to compute) that if there is a counterexample then $a>10.000.000$ (the above source is for 1 million, also note gmp uses the rabin miller test which only tells you if a number is probably prime but for prime numbers that are small like the ones in this program, it works correct [I think no example where it fails is known]).

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.