Currently there may be errors shown on top of a page, because of a missing Wiki update (PHP version and extension DPL3).
Navigation
Topics Help • Register • News • History • How to • Sequences statistics • Template prototypes

Difference between revisions of "Fermat number"

From Prime-Wiki
Jump to: navigation, search
(navbox)
m
 
(3 intermediate revisions by the same user not shown)
Line 1: Line 1:
 +
__TOC__
 
In [[mathematics]], a '''Fermat number''', named after [[Pierre de Fermat]] who first studied them, is a positive integer of the form
 
In [[mathematics]], a '''Fermat number''', named after [[Pierre de Fermat]] who first studied them, is a positive integer of the form
:<math>F_{n} = 2^{2^n} + 1</math>
+
:{{V|F}}<sub>{{Vn}}</sub> = {{Kbn|+|1|2|2<sup>n</sup>}}
where ''n'' is a [[non-negative]] [[integer]]. The first eight Fermat numbers are {{OEIS|A000215}}:
+
where {{Vn}} is a [[non-negative]] [[integer]]. The first eight Fermat numbers are (see {{OEIS|l|A000215}}):
  
:''F''<sub>0</sub> = 2<sup>1</sup> + 1 = 3
+
:{{V|F}}<sub>0</sub> = {{Kbn|+|1}} = 3
:''F''<sub>1</sub> = 2<sup>2</sup> + 1 = 5
+
:{{V|F}}<sub>1</sub> = {{Kbn|+|2}} = 5
:''F''<sub>2</sub> = 2<sup>4</sup> + 1 = 17
+
:{{V|F}}<sub>2</sub> = {{Kbn|+|4}} = 17
:''F''<sub>3</sub> = 2<sup>8</sup> + 1 = 257
+
:{{V|F}}<sub>3</sub> = {{Kbn|+|8}} = 257
:''F''<sub>4</sub> = 2<sup>16</sup> + 1 = 65537
+
:{{V|F}}<sub>4</sub> = {{Kbn|+|16}} = 65537
:''F''<sub>5</sub> = 2<sup>32</sup> + 1 = 4294967297 = 641 &times; 6700417
+
:{{V|F}}<sub>5</sub> = {{Kbn|+|32}} = 4294967297 = 641 &times; 6700417
:''F''<sub>6</sub> = 2<sup>64</sup> + 1 = 18446744073709551617 = 274177 &times; 67280421310721
+
:{{V|F}}<sub>6</sub> = {{Kbn|+|64}} = 18446744073709551617 = 274177 &times; 67280421310721
:''F''<sub>7</sub> = 2<sup>128</sup> + 1 = 340282366920938463463374607431768211457 = 59649589127497217 &times; 5704689200685129054721
+
:{{V|F}}<sub>7</sub> = {{Kbn|+|128}} = 340282366920938463463374607431768211457 = 59649589127497217 &times; 5704689200685129054721
  
 
Only the first 12 Fermat numbers have been completely factored. These factorisations can be found at [http://www.prothsearch.com/fermat.html Prime Factors of Fermat Numbers]
 
Only the first 12 Fermat numbers have been completely factored. These factorisations can be found at [http://www.prothsearch.com/fermat.html Prime Factors of Fermat Numbers]
  
If ''n'' is a positive integer and 2<sup>''n''</sup> + 1 is [[prime]], it can be shown that ''n'' must be a power of 2. (If ''n'' = ''ab'' where 1 < ''a'', ''b'' < ''n'' and ''b'' is odd, then 2<sup>''n''</sup> + 1 &equiv; (2<sup>''a''</sup>)<sup>''b''</sup> + 1 &equiv; (&minus;1)<sup>''b''</sup> + 1 &equiv; 0 (mod 2<sup>''a''</sup> + 1).) In other words, every prime of the form 2<sup>''n''</sup> + 1 is a Fermat number, and such primes are called '''Fermat primes'''. The only known Fermat primes are ''F''<sub>0</sub>,...,''F''<sub>4</sub>.
+
If {{Vn}} is a positive integer and {{Kbn|+|n}} is [[prime]], it can be shown that {{Vn}} must be a power of 2. (If {{Vn}} = {{V|ab}} where 1 < {{V|a}}, {{V|b}} < {{Vn}} and {{V|b}} is odd, then {{Kbn|+|n}} &equiv; (2<sup>{{V|a}}</sup>)<sup>{{V|b}}</sup> + 1 &equiv; (-1)<sup>{{V|b}}</sup> + 1 &equiv; 0 (mod 2<sup>{{V|a}}</sup> + 1).) In other words, every prime of the form {{Kbn|+|n}} is a Fermat number, and such primes are called '''Fermat primes'''. The only known Fermat primes are {{V|F}}<sub>0</sub>,...,{{V|F}}<sub>4</sub>.
  
 
==Basic properties==
 
==Basic properties==
Line 23: Line 24:
 
:<math>F_{n} = F_{0} \cdots F_{n-1} + 2</math>
 
:<math>F_{n} = F_{0} \cdots F_{n-1} + 2</math>
  
for ''n'' &ge; 2. Each of these relations can be proved by [[Proof by induction|mathematical induction]]. From the last equation, we can deduce '''Goldbach's theorem''': no two Fermat numbers share a common factor. To see this, suppose that 0 &le; ''i'' < ''j'' and ''F''<sub>''i''</sub> and ''F''<sub>''j''</sub> have a common factor ''a'' > 1. Then ''a'' divides both
+
for {{Vn}} &ge; 2. Each of these relations can be proved by [[Proof by induction|mathematical induction]]. From the last equation, we can deduce '''Goldbach's theorem''': no two Fermat numbers share a common factor. To see this, suppose that 0 &le; {{V|i}} < {{V|j}} and {{V|F}}<sub>{{V|i}}</sub> and {{V|F}}<sub>{{V|j}}</sub> have a common factor {{V|a}} > 1. Then {{V|a}} divides both
 
:<math>F_{0} \cdots F_{j-1}</math>
 
:<math>F_{0} \cdots F_{j-1}</math>
and ''F''<sub>''j''</sub>; hence ''a'' divides their difference 2. Since ''a'' > 1, this forces ''a'' = 2. This is a contradiction, because each Fermat number is clearly odd. As a corollary, we obtain another proof of the infinitude of the prime numbers: for each ''F''<sub>''n''</sub>, choose a prime factor ''p''<sub>''n''</sub>; then the sequence {''p''<sub>''n''</sub>} is an infinite sequence of distinct primes.
+
and {{V|F}}<sub>{{V|j}}</sub>; hence {{V|a}} divides their difference 2. Since {{V|a}} > 1, this forces {{V|a}} = 2. This is a contradiction, because each Fermat number is clearly odd. As a corollary, we obtain another proof of the infinitude of the prime numbers: for each {{V|F}}<sub>{{Vn}}</sub>, choose a prime factor {{V|p}}<sub>{{Vn}}</sub>; then the sequence {{{V|p}}<sub>{{Vn}}</sub>} is an infinite sequence of distinct primes.
  
 
Here are some other basic properties of the Fermat numbers:
 
Here are some other basic properties of the Fermat numbers:
*If ''n'' &ge; 2, then ''F''<sub>''n''</sub> &equiv; 17 or 41 ('''mod''' 72). (See [[modular arithmetic]])
+
*If {{Vn}} &ge; 2, then {{V|F}}<sub>{{Vn}}</sub> &equiv; 17 or 41 ('''mod''' 72). (See [[modular arithmetic]])
*If ''n'' &ge; 2, then ''F''<sub>''n''</sub> &equiv; 17, 37, 57, or 97 ('''mod''' 100).
+
*If {{Vn}} &ge; 2, then {{V|F}}<sub>{{Vn}}</sub> &equiv; 17, 37, 57, or 97 ('''mod''' 100).
*The number of digits ''D''(''n'',''b'') of ''F''<sub>''n''</sub> expressed in the base ''b'' is
+
*The number of digits {{V|D}}({{Vn}},{{Vb}}) of {{V|F}}<sub>{{Vn}}</sub> expressed in the base {{Vb}} is
 
:<math>D(n,b) = \lfloor \log_{b}\left(2^{2^{n}}+1\right)+1 \rfloor \approx \lfloor 2^{n}\,\log_{b}2+1 \rfloor </math> (See [[floor function]])
 
:<math>D(n,b) = \lfloor \log_{b}\left(2^{2^{n}}+1\right)+1 \rfloor \approx \lfloor 2^{n}\,\log_{b}2+1 \rfloor </math> (See [[floor function]])
 
*No Fermat number can be expressed as the sum of two primes, with the exception of F<sub>1</sub> = 2 + 3.
 
*No Fermat number can be expressed as the sum of two primes, with the exception of F<sub>1</sub> = 2 + 3.
Line 36: Line 37:
  
 
==Primality of Fermat numbers==
 
==Primality of Fermat numbers==
Fermat numbers and Fermat primes were first studied by Pierre de Fermat, who conjectured that all Fermat numbers are prime. Indeed, the first five Fermat numbers ''F''<sub>0</sub>,...,''F''<sub>4</sub> are easily shown to be prime. However, this conjecture was refuted by [[Leonhard Euler]] in 1732 when he showed that  
+
Fermat numbers and Fermat primes were first studied by Pierre de Fermat, who conjectured that all Fermat numbers are prime. Indeed, the first five Fermat numbers {{V|F}}<sub>0</sub>,...,{{V|F}}<sub>4</sub> are easily shown to be prime. However, this conjecture was refuted by [[Leonhard Euler]] in 1732 when he showed that  
 
:<math>F_{5} = 2^{2^5} + 1 = 2^{32} + 1 = 4294967297 = 641 \cdot 6700417</math>
 
:<math>F_{5} = 2^{2^5} + 1 = 2^{32} + 1 = 4294967297 = 641 \cdot 6700417</math>
 
*See [http://math.dartmouth.edu/~euler/pages/E026.html Euler's work and Fermat's two letters]. Original document and English translation is available.  
 
*See [http://math.dartmouth.edu/~euler/pages/E026.html Euler's work and Fermat's two letters]. Original document and English translation is available.  
  
It is interesting to note how Euler found this [[factorization]]. Euler had proved that every factor of ''F''<sub>''n''</sub> must have the form ''k''2<sup>''n''+1</sup> + 1. For ''n'' = 5, this means that the only possible factors are of the form 64''k'' + 1. It did not take Euler very long to find the factor 641 = 10&times;64 + 1. It isn't hard to check this, but there is an interesting proof for this: <math>641=2^4+5^4=5*2^7+1</math> using this we can get
+
It is interesting to note how Euler found this [[factorization]]. Euler had proved that every factor of {{V|F}}<sub>{{Vn}}</sub> must have the form {{Vk}}{{Kbn|+|n+1}}. For {{Vn}} = 5, this means that the only possible factors are of the form 64{{Vk}} + 1. It did not take Euler very long to find the factor 641 = 10&times;64 + 1. It isn't hard to check this, but there is an interesting proof for this: <math>641=2^4+5^4=5*2^7+1</math> using this we can get
:<math>GCD(2^{2^5}+1,641)=GCD(2^{4}*2^{28}+1,641)=GCD(-5^{4}*2^{28}+1,641)=GCD((5*2^{7})^{4}-1,641)=GCD((-1)^{4}-1,641)=GCD(0,641)=641</math>
+
:<math>GCD(2^{2^5}+1,641)=GCD(2^{4}*2^{28}+1,641)=GCD(-5^{4}*2^{28}+1,641)=GCD((5*2^{7})^{4}-1,641)\\=GCD((-1)^{4}-1,641)=GCD(0,641)=641</math>
So <math>F_{5}</math> is divisible by 641.
+
So {{V|F}}<sub>5</sub> is divisible by 641.
  
 
It is widely believed that Fermat was aware of Euler's result, so it seems curious why he failed to follow through on the straightforward calculation to find the factor. One common explanation is that Fermat made a computational mistake and was so convinced of the correctness of his claim that he failed to double-check his work.
 
It is widely believed that Fermat was aware of Euler's result, so it seems curious why he failed to follow through on the straightforward calculation to find the factor. One common explanation is that Fermat made a computational mistake and was so convinced of the correctness of his claim that he failed to double-check his work.
  
There are no other known Fermat primes ''F''<sub>''n''</sub> with ''n'' > 4. In fact, each of the following is an open problem:
+
There are no other known Fermat primes {{V|F}}<sub>{{Vn}}</sub> with {{Vn}} > 4. In fact, each of the following is an open problem:
*Is ''F''<sub>''n''</sub> composite for all ''n'' > 4?
+
*Is {{V|F}}<sub>{{Vn}}</sub> composite for all {{Vn}} > 4?
 
*Are there infinitely many Fermat primes?
 
*Are there infinitely many Fermat primes?
 
*Are there infinitely many composite Fermat numbers?
 
*Are there infinitely many composite Fermat numbers?
*Is every Fermat numbers squarefree?
+
*Is every Fermat number squarefree?
  
The following heuristic argument suggests there are only finitely many Fermat primes: according to the [[prime number theorem]], the "probability" that a random number ''n'' is prime is at most ''A''/ln(''n''), where ''A'' is a fixed constant. Therefore, the total expected number of Fermat primes is at most
+
The following heuristic argument suggests there are only finitely many Fermat primes: according to the [[prime number theorem]], the "probability" that a random number {{Vn}} is prime is at most {{V|A}}/ln({{Vn}}), where {{V|A}} is a fixed constant. Therefore, the total expected number of Fermat primes is at most
 
:<math>A \sum_{n=0}^{\infty} \frac{1}{\ln F_{n}} = \frac{A}{\ln 2} \sum_{n=0}^{\infty} \frac{1}{\log_{2}(2^{2^{n}}+1)} < \frac{A}{\ln 2} \sum_{n=0}^{\infty} 2^{-n} = \frac{2A}{\ln 2}</math>
 
:<math>A \sum_{n=0}^{\infty} \frac{1}{\ln F_{n}} = \frac{A}{\ln 2} \sum_{n=0}^{\infty} \frac{1}{\log_{2}(2^{2^{n}}+1)} < \frac{A}{\ln 2} \sum_{n=0}^{\infty} 2^{-n} = \frac{2A}{\ln 2}</math>
  
It should be stressed that this argument is in no way a rigorous proof. For one thing, the argument assumes that Fermat numbers behave "randomly", yet we have already seen that the factors of Fermat numbers have special properties. Although it is widely believed that there are only finitely many Fermat primes, it should be noted that there are some experts who disagree. [http://www.spd.dcu.ie/johnbcos/fermat6.htm]
+
It should be stressed that this argument is in no way a rigorous proof. For one thing, the argument assumes that Fermat numbers behave "randomly", yet we have already seen that the factors of Fermat numbers have special properties. Although it is widely believed that there are only finitely many Fermat primes, it should be noted that there are some experts who disagree ([https://johnbcosgrave.com/archive/fermat6.htm John Cosgrave: "Fermat 6"]).
  
As of this writing (2004), it is known that ''F''<sub>''n''</sub> is composite for 5 &le; ''n'' &le; 32, although complete factorisations of ''F''<sub>''n''</sub> are known only for 0 &le; ''n'' &le; 11. The largest known composite Fermat number is ''F''<sub>2478782</sub>, and its prime factor 3&times;2<sup>2478785</sup> + 1 was discovered by John Cosgrave and his Proth-Gallot Group on October 10, 2003.
+
As of this writing (2004), it is known that {{V|F}}<sub>{{Vn}}</sub> is composite for 5 &le; {{Vn}} &le; 32, although complete factorisations of {{V|F}}<sub>{{Vn}}</sub> are known only for 0 &le; {{Vn}} &le; 11.
  
There are a number of conditions that are equivalent to the primality of ''F''<sub>''n''</sub>.
+
The largest known composite Fermat number is {{NGF|18233954}}, and its prime factor {{NPr|7|18233956}} was discovered by [[Ryan Propper]] on {{T5000|131289|2020-10-01}}.
  
*'''[[Proth's theorem]]''' -- (1878) Let ''N'' = ''k''2<sup>''m''</sup> + 1 with odd ''k'' < 2<sup>''m''</sup>. If there is an integer ''a'' such that
+
There are a number of conditions that are equivalent to the primality of {{V|F}}<sub>{{Vn}}</sub>.
 +
 
 +
*'''[[Proth's theorem]]''' -- (1878) Let {{V|N}} = {{Kbn|+|k|2|m}} with odd {{Vk}} < 2<sup>{{V|m}}</sup>. If there is an integer {{V|a}} such that
 
:<math>a^{(N-1)/2} \equiv -1 \,\pmod N </math>
 
:<math>a^{(N-1)/2} \equiv -1 \,\pmod N </math>
:then ''N'' is prime. Conversely, if the above congruence does not hold, and in addition
+
:then {{V|N}} is prime. Conversely, if the above congruence does not hold, and in addition
:<math>\left(\frac{a}{N}\right)=-1</math> (See [[Wikipedia:Jacobi symbol]])
+
:<math>\left(\frac{a}{N}\right)=-1</math> (See [[Wikipedia:Jacobi symbol|Jacobi symbol]])
  
:then ''N'' is composite. If ''N'' = ''F''<sub>''n''</sub> > 3, then the above Jacobi symbol is always equal to &minus;1 for ''a'' = 3, and this special case of Proth's theorem is known as '''[[Pépin's test]]'''. Although Pépin's test and Proth's theorem have been implemented on computers to prove the compositeness of many Fermat numbers, neither test gives a specific nontrivial factor. In fact, no specific prime factors are known for ''n'' = 14, 20, 22, and 24.
+
:then {{V|N}} is composite. If {{V|N}} = {{V|F}}<sub>{{Vn}}</sub> > 3, then the above Jacobi symbol is always equal to &minus;1 for {{V|a}} = 3, and this special case of Proth's theorem is known as '''[[Pépin's test]]'''. Although Pépin's test and Proth's theorem have been implemented on computers to prove the compositeness of many Fermat numbers, neither test gives a specific nontrivial factor. In fact, no specific prime factors are known for {{Vn}} = 14, 20, 22, and 24.
* Theorem: Let n>0 then <math>F_n</math> is prime if and only if it divides <math>S_{2^n-2}</math> where <math>S_0=5</math> and <math>S_{k+1}=S^2_{k}-2</math> [ http://www.robertgerbicz.tar.hu/Fermatnumbers.pdf See the proof by Robert Gerbicz]
+
* Theorem: Let {{Vn}}>0 then <math>F_n</math> is prime if and only if it divides <math>S_{2^n-2}</math> where <math>S_0=5</math> and <math>S_{k+1}=S^2_{k}-2</math> ([http://www.robertgerbicz.tar.hu/Fermatnumbers.pdf Proof] by Robert Gerbicz).
*Let ''n'' &ge; 3 be a positive odd integer. Then ''n'' is a Fermat prime if and only if for every ''a'' coprime to ''n'', ''a'' is a [[primitive root modulo n|primitive root]] mod ''n'' if and only if ''a'' is a [[Quadratic residue|quadratic nonresidue]] mod ''n''.
+
*Let {{Vn}} &ge; 3 be a positive odd integer. Then {{Vn}} is a Fermat prime if and only if for every {{V|a}} coprime to {{Vn}}, {{V|a}} is a [[primitive root modulo n|primitive root]] mod {{V|n}} if and only if {{V|a}} is a [[Quadratic residue|quadratic nonresidue]] mod {{Vn}}.
*The Fermat number ''F''<sub>''n''</sub> > 3 is prime if and only if it can be written uniquely as a sum of two nonzero squares, namely
+
*The Fermat number {{V|F}}<sub>{{Vn}}</sub> > 3 is prime if and only if it can be written uniquely as a sum of two nonzero squares, namely
 
:<math>F_{n}=\left(2^{2^{n-1}}\right)^{2}+1^{2}</math>
 
:<math>F_{n}=\left(2^{2^{n-1}}\right)^{2}+1^{2}</math>
 
:When <math>F_{n} = x^2 + y^2</math> not of the form shown above, a proper factor is:
 
:When <math>F_{n} = x^2 + y^2</math> not of the form shown above, a proper factor is:
 
:<math>\gcd(x + 2^{2^{n-1}} y, F_{n})</math>
 
:<math>\gcd(x + 2^{2^{n-1}} y, F_{n})</math>
:Example 1: ''F''<sub>5</sub> = 62264<sup>2</sup> + 20449<sup>2</sup>, so a proper factor is <math>\gcd(62264\, +\, 2^{2^4}\, 20449,\, F_{5}) = 641</math>.
+
:Example 1: {{V|F}}<sub>5</sub> = 62264<sup>2</sup> + 20449<sup>2</sup>, so a proper factor is <math>\gcd(62264\, +\, 2^{2^4}\, 20449,\, F_{5}) = 641</math>.
:Example 2: ''F''<sub>6</sub> = 4046803256<sup>2</sup> + 1438793759<sup>2</sup>, so a proper factor is <math>\gcd(4046803256\, +\, 2^{2^5}\, 1438793759,\, F_{6}) = 274177</math>.
+
:Example 2: {{V|F}}<sub>6</sub> = 4046803256<sup>2</sup> + 1438793759<sup>2</sup>, so a proper factor is <math>\gcd(4046803256\, +\, 2^{2^5}\, 1438793759,\, F_{6}) = 274177</math>.
  
 
==Factorization of Fermat numbers==
 
==Factorization of Fermat numbers==
Because or the size of Fermat numbers, it is difficult to factorize or to prove primality of those. Pépin's test is necessary and sufficient test for primality of Fermat numbers which can be implemented by modern computers. Elliptic curve method is a fast method for finding small prime divisors of numbers, and at least [[GIMPS]] is trying to find prime divisors of Fermat numbers by elliptic curve method. Distributed computing project Fermatsearch has also succesfully found some factors of Fermat numbers. Yves Gallot's proth.exe has been used to find factors of large Fermat numbers. Lucas proved in year 1878 that every factor of Fermat number <math>F_n</math> is of the form <math>k2^{n+2}+1</math>, where k is a nonnegative integer.
+
Because or the size of Fermat numbers, it is difficult to factorize or to prove primality of those. Pépin's test is necessary and sufficient test for primality of Fermat numbers which can be implemented by modern computers. Elliptic curve method is a fast method for finding small prime divisors of numbers, and at least [[GIMPS]] is trying to find prime divisors of Fermat numbers by elliptic curve method. Distributed computing project [[FermatSearch]] has also succesfully found some factors of Fermat numbers. Yves Gallot's proth.exe has been used to find factors of large Fermat numbers. Lucas proved in year 1878 that every factor of Fermat number <math>F_n</math> is of the form <math>k2^{n+2}+1</math>, where k is a nonnegative integer.
  
 
* [http://wwwmaths.anu.edu.au/~brent/pub/pub061.html Factorization of the eighth Fermat number]
 
* [http://wwwmaths.anu.edu.au/~brent/pub/pub061.html Factorization of the eighth Fermat number]
Line 89: Line 92:
  
 
==Relationship to constructible polygons==
 
==Relationship to constructible polygons==
An ''n''-sided regular polygon can be constructed with [[Wikipedia:ruler and compass|ruler and compass]] if and only if ''n'' is a power of 2 or the product of a power of 2 and distinct Fermat primes. In other words, if and only if ''n'' is of the form ''n'' = 2<sup>''k''</sup>''p''<sub>1</sub>''p''<sub>2</sub>...''p''<sub>''s''</sub>, where ''k'' is a nonnegative integer and the ''p''<sub>''i''</sub> are distinct Fermat primes. See [[Wikipedia:constructible polygon|constructible polygon]].
+
An {{Vn}}-sided regular polygon can be constructed with [[Wikipedia:ruler and compass|ruler and compass]] if and only if {{Vn}} is a power of 2 or the product of a power of 2 and distinct Fermat primes. In other words, if and only if {{Vn}} is of the form {{Vn}} = 2<sup>{{Vk}}</sup>{{V|p}}<sub>1</sub>{{V|p}}<sub>2</sub>...{{V|p}}<sub>{{V|s}}</sub>, where {{Vk}} is a nonnegative integer and the {{V|p}}<sub>{{V|i}}</sub> are distinct Fermat primes. See [[Wikipedia:constructible polygon|constructible polygon]].
  
A positive integer ''n'' is of the above form if and only if &phi;(''n'') is a power of 2, where &phi;(''n'') is [[Euler's totient function]].
+
A positive integer {{Vn}} is of the above form if and only if &phi;({{Vn}}) is a power of 2, where &phi;({{Vn}}) is [[Euler's totient function]].
  
 
==Applications of Fermat numbers==
 
==Applications of Fermat numbers==
Line 97: Line 100:
  
 
==Other interesting facts==
 
==Other interesting facts==
...F<sub>n</sub> cannot be a perfect power, perfect, or part of amicable pair, etc...
+
...{{V|F}}<sub>{{Vn}}</sub> cannot be a perfect power, perfect, or part of amicable pair, etc...
  
 
==Generalized Fermat numbers==
 
==Generalized Fermat numbers==
Line 124: Line 127:
 
* ''17 Lectures on Fermat Numbers: From Number Theory to Geometry'', Michal Krizek, Florian Luca, Lawrence Somer, Springer, CMS Books 9, ISBN 0387953329 (This book contains an extensive list of references.)
 
* ''17 Lectures on Fermat Numbers: From Number Theory to Geometry'', Michal Krizek, Florian Luca, Lawrence Somer, Springer, CMS Books 9, ISBN 0387953329 (This book contains an extensive list of references.)
 
{{Navbox NumberClasses}}
 
{{Navbox NumberClasses}}
[[Category:Number]]
+
[[Category:Generalized Fermat number| ]]

Latest revision as of 14:35, 9 August 2021

In mathematics, a Fermat number, named after Pierre de Fermat who first studied them, is a positive integer of the form

Fn = 22n+1

where n is a non-negative integer. The first eight Fermat numbers are (see sequence A000215 in OEIS):

F0 = 21+1 = 3
F1 = 22+1 = 5
F2 = 24+1 = 17
F3 = 28+1 = 257
F4 = 216+1 = 65537
F5 = 232+1 = 4294967297 = 641 × 6700417
F6 = 264+1 = 18446744073709551617 = 274177 × 67280421310721
F7 = 2128+1 = 340282366920938463463374607431768211457 = 59649589127497217 × 5704689200685129054721

Only the first 12 Fermat numbers have been completely factored. These factorisations can be found at Prime Factors of Fermat Numbers

If n is a positive integer and 2n+1 is prime, it can be shown that n must be a power of 2. (If n = ab where 1 < a, b < n and b is odd, then 2n+1 ≡ (2a)b + 1 ≡ (-1)b + 1 ≡ 0 (mod 2a + 1).) In other words, every prime of the form 2n+1 is a Fermat number, and such primes are called Fermat primes. The only known Fermat primes are F0,...,F4.

Basic properties

The Fermat numbers satisfy the following recurrence relations

[math]\displaystyle{ F_{n} = (F_{n-1}-1)^{2}+1 }[/math]
[math]\displaystyle{ F_{n} = F_{n-1} + 2^{2^{n-1}}F_{0} \cdots F_{n-2} }[/math]
[math]\displaystyle{ F_{n} = F_{n-1}^2 - 2(F_{n-2}-1)^2 }[/math]
[math]\displaystyle{ F_{n} = F_{0} \cdots F_{n-1} + 2 }[/math]

for n ≥ 2. Each of these relations can be proved by mathematical induction. From the last equation, we can deduce Goldbach's theorem: no two Fermat numbers share a common factor. To see this, suppose that 0 ≤ i < j and Fi and Fj have a common factor a > 1. Then a divides both

[math]\displaystyle{ F_{0} \cdots F_{j-1} }[/math]

and Fj; hence a divides their difference 2. Since a > 1, this forces a = 2. This is a contradiction, because each Fermat number is clearly odd. As a corollary, we obtain another proof of the infinitude of the prime numbers: for each Fn, choose a prime factor pn; then the sequence {pn} is an infinite sequence of distinct primes.

Here are some other basic properties of the Fermat numbers:

  • If n ≥ 2, then Fn ≡ 17 or 41 (mod 72). (See modular arithmetic)
  • If n ≥ 2, then Fn ≡ 17, 37, 57, or 97 (mod 100).
  • The number of digits D(n,b) of Fn expressed in the base b is
[math]\displaystyle{ D(n,b) = \lfloor \log_{b}\left(2^{2^{n}}+1\right)+1 \rfloor \approx \lfloor 2^{n}\,\log_{b}2+1 \rfloor }[/math] (See floor function)
  • No Fermat number can be expressed as the sum of two primes, with the exception of F1 = 2 + 3.
  • No Fermat prime can be expressed as the difference of two pth powers, where p is an odd prime.

Primality of Fermat numbers

Fermat numbers and Fermat primes were first studied by Pierre de Fermat, who conjectured that all Fermat numbers are prime. Indeed, the first five Fermat numbers F0,...,F4 are easily shown to be prime. However, this conjecture was refuted by Leonhard Euler in 1732 when he showed that

[math]\displaystyle{ F_{5} = 2^{2^5} + 1 = 2^{32} + 1 = 4294967297 = 641 \cdot 6700417 }[/math]

It is interesting to note how Euler found this factorization. Euler had proved that every factor of Fn must have the form k2n+1+1. For n = 5, this means that the only possible factors are of the form 64k + 1. It did not take Euler very long to find the factor 641 = 10×64 + 1. It isn't hard to check this, but there is an interesting proof for this: [math]\displaystyle{ 641=2^4+5^4=5*2^7+1 }[/math] using this we can get

[math]\displaystyle{ GCD(2^{2^5}+1,641)=GCD(2^{4}*2^{28}+1,641)=GCD(-5^{4}*2^{28}+1,641)=GCD((5*2^{7})^{4}-1,641)\\=GCD((-1)^{4}-1,641)=GCD(0,641)=641 }[/math]

So F5 is divisible by 641.

It is widely believed that Fermat was aware of Euler's result, so it seems curious why he failed to follow through on the straightforward calculation to find the factor. One common explanation is that Fermat made a computational mistake and was so convinced of the correctness of his claim that he failed to double-check his work.

There are no other known Fermat primes Fn with n > 4. In fact, each of the following is an open problem:

  • Is Fn composite for all n > 4?
  • Are there infinitely many Fermat primes?
  • Are there infinitely many composite Fermat numbers?
  • Is every Fermat number squarefree?

The following heuristic argument suggests there are only finitely many Fermat primes: according to the prime number theorem, the "probability" that a random number n is prime is at most A/ln(n), where A is a fixed constant. Therefore, the total expected number of Fermat primes is at most

[math]\displaystyle{ A \sum_{n=0}^{\infty} \frac{1}{\ln F_{n}} = \frac{A}{\ln 2} \sum_{n=0}^{\infty} \frac{1}{\log_{2}(2^{2^{n}}+1)} \lt \frac{A}{\ln 2} \sum_{n=0}^{\infty} 2^{-n} = \frac{2A}{\ln 2} }[/math]

It should be stressed that this argument is in no way a rigorous proof. For one thing, the argument assumes that Fermat numbers behave "randomly", yet we have already seen that the factors of Fermat numbers have special properties. Although it is widely believed that there are only finitely many Fermat primes, it should be noted that there are some experts who disagree (John Cosgrave: "Fermat 6").

As of this writing (2004), it is known that Fn is composite for 5 ≤ n ≤ 32, although complete factorisations of Fn are known only for 0 ≤ n ≤ 11.

The largest known composite Fermat number is Fermat 2218233954+1, and its prime factor Proth 7•218233956+1 was discovered by Ryan Propper on 2020-10-01.

There are a number of conditions that are equivalent to the primality of Fn.

  • Proth's theorem -- (1878) Let N = k•2m+1 with odd k < 2m. If there is an integer a such that
[math]\displaystyle{ a^{(N-1)/2} \equiv -1 \,\pmod N }[/math]
then N is prime. Conversely, if the above congruence does not hold, and in addition
[math]\displaystyle{ \left(\frac{a}{N}\right)=-1 }[/math] (See Jacobi symbol)
then N is composite. If N = Fn > 3, then the above Jacobi symbol is always equal to −1 for a = 3, and this special case of Proth's theorem is known as Pépin's test. Although Pépin's test and Proth's theorem have been implemented on computers to prove the compositeness of many Fermat numbers, neither test gives a specific nontrivial factor. In fact, no specific prime factors are known for n = 14, 20, 22, and 24.
  • Theorem: Let n>0 then [math]\displaystyle{ F_n }[/math] is prime if and only if it divides [math]\displaystyle{ S_{2^n-2} }[/math] where [math]\displaystyle{ S_0=5 }[/math] and [math]\displaystyle{ S_{k+1}=S^2_{k}-2 }[/math] (Proof by Robert Gerbicz).
  • Let n ≥ 3 be a positive odd integer. Then n is a Fermat prime if and only if for every a coprime to n, a is a primitive root mod n if and only if a is a quadratic nonresidue mod n.
  • The Fermat number Fn > 3 is prime if and only if it can be written uniquely as a sum of two nonzero squares, namely
[math]\displaystyle{ F_{n}=\left(2^{2^{n-1}}\right)^{2}+1^{2} }[/math]
When [math]\displaystyle{ F_{n} = x^2 + y^2 }[/math] not of the form shown above, a proper factor is:
[math]\displaystyle{ \gcd(x + 2^{2^{n-1}} y, F_{n}) }[/math]
Example 1: F5 = 622642 + 204492, so a proper factor is [math]\displaystyle{ \gcd(62264\, +\, 2^{2^4}\, 20449,\, F_{5}) = 641 }[/math].
Example 2: F6 = 40468032562 + 14387937592, so a proper factor is [math]\displaystyle{ \gcd(4046803256\, +\, 2^{2^5}\, 1438793759,\, F_{6}) = 274177 }[/math].

Factorization of Fermat numbers

Because or the size of Fermat numbers, it is difficult to factorize or to prove primality of those. Pépin's test is necessary and sufficient test for primality of Fermat numbers which can be implemented by modern computers. Elliptic curve method is a fast method for finding small prime divisors of numbers, and at least GIMPS is trying to find prime divisors of Fermat numbers by elliptic curve method. Distributed computing project FermatSearch has also succesfully found some factors of Fermat numbers. Yves Gallot's proth.exe has been used to find factors of large Fermat numbers. Lucas proved in year 1878 that every factor of Fermat number [math]\displaystyle{ F_n }[/math] is of the form [math]\displaystyle{ k2^{n+2}+1 }[/math], where k is a nonnegative integer.

Fermat's little theorem and pseudoprimes

...Using Fermat numbers to generate infinitely many pseudoprimes...

Relationship to constructible polygons

An n-sided regular polygon can be constructed with ruler and compass if and only if n is a power of 2 or the product of a power of 2 and distinct Fermat primes. In other words, if and only if n is of the form n = 2kp1p2...ps, where k is a nonnegative integer and the pi are distinct Fermat primes. See constructible polygon.

A positive integer n is of the above form if and only if φ(n) is a power of 2, where φ(n) is Euler's totient function.

Applications of Fermat numbers

...Fermat number transform...random number generation...

Other interesting facts

...Fn cannot be a perfect power, perfect, or part of amicable pair, etc...

Generalized Fermat numbers

Main title: Generalized Fermat number ...brief definition of L(p,m) and G(p,m)...

See also

External links

References:

  • 17 Lectures on Fermat Numbers: From Number Theory to Geometry, Michal Krizek, Florian Luca, Lawrence Somer, Springer, CMS Books 9, ISBN 0387953329 (This book contains an extensive list of references.)
Number classes
General numbers
Special numbers
Prime numbers