Odd perfect numbers
Is there an odd number that equals the sum of its divisors other than itself?
Summary
Why it is open
A number N is perfect when σ(N) = 2N, where σ(N) is the sum of all positive divisors of N, N included [3]. An even number is perfect exactly when it has the form 2^(k−1)(2^k − 1) with 2^k − 1 a prime, called a Mersenne prime [3]. Whether an odd perfect number exists is, in the words of Pascal Ochem and Michaël Rao, "a long-standing open question" [3].
A resolution would be a single odd perfect number or a proof that none exists. What has accumulated instead is a long list of necessary conditions, properties that any odd perfect number would have to satisfy; the published bounds are theorems, and several rest on large computer searches [3, 8]. They narrow where an example could be, but none so far excludes every case [3, 8].
Pomerance offered a heuristic case against odd perfect numbers [10]. An argument communicated by Carl Pomerance makes the expected number of odd perfect numbers tiny, but the same reasoning would also forbid large even perfect numbers, which exist [10].
What is known
Euler's form and finiteness (1913–2015). Leonhard Euler proved that an odd perfect number has the form N = p^e m², where p is a prime with p ≡ e ≡ 1 (mod 4) and p does not divide m; p is called the special prime [4]. Write ω(N) for the number of distinct prime factors of N and Ω(N) for the number counted with multiplicity; Euler's form gives Ω(N) ≥ 2ω(N) − 1, and R. Steuerwald's 1937 result that the other primes cannot all appear exactly squared gives Ω(N) ≥ 2ω(N) + 1 [4]. In 1913 Leonard Eugene Dickson proved that there are only finitely many odd perfect numbers with k distinct prime factors; Pomerance gave an explicit bound in 1977, D. R. Heath-Brown improved it to N < 4^(4^k), R. J. Cook lowered it further, and Pace Nielsen proved N < 2^(4^k) in 2003 [1]. In 2015 Nielsen sharpened this to 10¹² P² N < 2^(4^k), where P is the largest prime factor of N [8].
Size (1991–2012). Richard Brent, Graeme Cohen and Herman te Riele proved in 1991 that N > 10³⁰⁰ [2]. Ochem and Rao extended the method in 2012 to N > 10¹⁵⁰⁰, and showed that N has at least 101 prime factors counted with multiplicity and that its largest component, a prime power dividing N exactly, exceeds 10⁶² [3]. The project's web page, read on 15 October 2026, states that the computation has since been pushed to N > 10²²⁰⁰ and Ω(N) ≥ 115, and gives a bound m > 10¹⁰⁰⁰ for Euler's form that it calls unpublished; no paper is linked for the first two [6].
Distinct prime factors (1979–2015). E. Z. Chein, in his doctoral thesis, and Peter Hagis independently proved that an odd perfect number has at least 8 distinct prime factors [1]. Nielsen raised this to 9 in 2007, and to 12 when 3 does not divide N [7]. In 2015 he proved that there are at least 10, after one computation that took a few months and a final search of a little over thirty million cases [8].
Prime factors counted with multiplicity (2014–2023). Ochem and Rao proved in 2014 that Ω(N) ≥ (18ω(N) − 31)/7 and Ω(N) ≥ 2ω(N) + 51 [4]. After improvements by Joshua Zelinsky, Graeme Clayton and Cody Hansen proved in 2023 that Ω(N) ≥ (99ω(N) − 187)/37, and Ω(N) ≥ (51ω(N) − 46)/19 when 3 does not divide N [5].
Individual prime factors (1999–2012). Takeshi Goto and Yasuo Ohno proved in 2008 that the largest prime factor exceeds 10⁸ [9]. Douglas Iannucci proved that the second largest exceeds 10⁴ and the third largest exceeds 100 [1]. In the other direction, Peter Acquaah and Sergei Konyagin proved in 2012 that every prime factor q satisfies q < (3N)^(1/3) [13].
Special forms (2026). A July 2026 preprint by Ochem and Zelinsky, not peer reviewed, treats odd perfect numbers in which every prime other than the special prime appears exactly squared except one. It shows that this exceptional prime must be 3 and that its exponent is at least 23,000,000,000, against a previous bound of 16 [12].
Main approaches
Computer proofs with factor chains. Assume in turn that each possible prime power divides N exactly, use the factors of the divisor sum to find further primes, and stop a branch when its abundancy σ(n)/n exceeds 2 or the number exceeds the target bound [3]. Strongest evidence: the 2012 bounds N > 10¹⁵⁰⁰, Ω(N) ≥ 101 and a component above 10⁶², with the programs made available [3]. Main difficulty: a branch reaches a "roadblock" when the only unexplored factors are composites that nobody has factored [3]; in 1991 Brent, Cohen and te Riele found that "numerous other 'unattainable' factorisations appeared to bar our way" [2], and the project's page lists the most difficult composites for a bound of 10²²⁰⁰ [6].
Counting distinct primes. Combine upper bounds in terms of k with computation to exclude small values of k [8]. Strongest evidence: k ≥ 10, obtained by turning the doubly exponential bound into "an effective estimation tool" [8]. Main difficulty: in a section titled "Can we prove k > 10?", Nielsen writes "The simple answer to the question in the title of this section is no", and describes a case whose next unknown prime lies between 1.2 × 10¹⁹ and 2.6 × 10¹⁹ [8].
Inequalities between Ω and ω. If most primes appeared only squared, their divisor sums would force some prime, such as 3, to appear to a high power; making this precise raises the lower bound for Ω in terms of ω [4, 5]. Strongest evidence: the slope 99/37, about 2.68, of Clayton and Hansen [5]. Main difficulty: these inequalities constrain an odd perfect number without contradicting its existence [4, 5].
Spoofs. Study solutions of the perfect-number equation in which the bases need not be primes [11]. René Descartes's example 3²·7²·11²·13²·22021 is spoof perfect, but 22021 = 19²·61 is not prime; John Voight found 3⁴·7²·11²·19²·(−127) [11]. Strongest evidence: the BYU group found all 21 nontrivial, odd, primitive spoofs with fewer than seven bases and proved that each number of bases allows only finitely many [11]. They conclude that approaches using only the multiplicative nature of σ "are unworkable", and that a proof must use both the primality and the positivity of the bases [11]. Main difficulty: spoofs show what a proof must use, not how to use it; the group found no example with only positive prime bases, and lists finiteness for non-primitive spoofs as open [11].
Heuristics. Treat divisibility as random [10]. In Pomerance's version, each odd m allows at most about log m choices of the prime in N = pm² (here p may divide m), each succeeding with "probability" 1/m², and the sum of (log m)/m² over m > 10⁷⁵ is below 10⁻⁷⁰ [10]. Strongest evidence: the convergence of that sum, which Ochem's unpublished bound m > 10¹⁰⁰⁰ is "tailor-made to boost" [6, 10]. Main difficulty: the heuristic also predicts no large even perfect numbers; Pomerance attributes the even case to the special structure of Mersenne primes and writes, "That's the trouble with heuristic arguments. They are not proofs" [10].
Open sub-questions
- Publishing larger bounds. Can the reported computations N > 10²²⁰⁰ and Ω(N) ≥ 115 be reproduced independently from the 2012 method and released with a proof tree that others can check, and which listed composites must be factored first [3, 6]?
- Eleven distinct primes. Nielsen says his method would need bounds up to 2.6 × 10¹⁹ for at least two of the primes 5, 61 and 163 to shrink the interval in his example; would that close the case, and which cases would then block a proof of k ≥ 11 [8]?
- The slope 3. Can the counting method behind Ω(N) ≥ (99ω(N) − 187)/37 reach a slope of 3 or more, and if not, what is the largest slope its constraints allow [5]?
- Spoofs with prime bases. Does an odd spoof perfect factorisation with only positive prime bases, repetitions allowed, exist with seven or eight bases [11]?
- Two exceptional exponents. Does the method of Ochem and Zelinsky give a comparable lower bound when exactly two primes other than the special prime appear to a power higher than 2 [12]?
- The largest prime factor. Can the cyclotomic method of Goto and Ohno raise the bound on the largest prime factor from 10⁸ to 10⁹, and at what computing cost [9]?
Work on this in Papex
If you have a testable answer to one of the sub-questions above, you can draft it as a private hypothesis in Papex, or prepare the draft from Claude or ChatGPT with the Papex connector. You review the exact version, sign it and decide whether to publish it; a public record that names this problem is listed on this page. A Papex record shows who signed which version and when. It is not peer review and does not mean the hypothesis is correct.
Key references
- Nielsen, P. P. (2003). An upper bound for odd perfect numbers. Integers 3, A14. Source · Checked 15 Oct 2026
- Brent, R. P., Cohen, G. L. & te Riele, H. J. J. (1991). Improved techniques for lower bounds for odd perfect numbers. Mathematics of Computation 57(196), 857–868. 10.1090/S0025-5718-1991-1094940-3 · Checked 15 Oct 2026
- Ochem, P. & Rao, M. (2012). Odd perfect numbers are greater than 10¹⁵⁰⁰. Mathematics of Computation 81(279), 1869–1877. 10.1090/S0025-5718-2012-02563-4 · Checked 15 Oct 2026
- Ochem, P. & Rao, M. (2014). On the number of prime factors of an odd perfect number. Mathematics of Computation 83(289), 2435–2439 (published online 2013). 10.1090/S0025-5718-2013-02776-7 · Checked 15 Oct 2026
- Clayton, G. & Hansen, C. (2023). On inequalities involving counts of the prime factors of an odd perfect number. Integers 23, A79. Source · Checked 15 Oct 2026
- Ochem, P. Odd perfect numbers. Web page of the Ochem–Rao project, LIRMM, Montpellier, with lists of composites dated February 2025. Source · Checked 15 Oct 2026
- Nielsen, P. P. (2007). Odd perfect numbers have at least nine distinct prime factors. Mathematics of Computation 76(260), 2109–2126. 10.1090/S0025-5718-07-01990-4 · Checked 15 Oct 2026
- Nielsen, P. P. (2015). Odd perfect numbers, Diophantine equations, and upper bounds. Mathematics of Computation 84(295), 2549–2567. 10.1090/S0025-5718-2015-02941-X · Checked 15 Oct 2026
- Goto, T. & Ohno, Y. (2008). Odd perfect numbers have a prime factor exceeding 10⁸. Mathematics of Computation 77(263), 1859–1868. 10.1090/S0025-5718-08-02050-4 · Checked 15 Oct 2026
- Lipp, W. Pomerance's heuristic that odd perfect numbers are unlikely. OddPerfect.org, quoting messages from Carl Pomerance; archived copy of 31 December 2006. Source · Checked 15 Oct 2026
- Andersen, N., Durham, S., Griffin, M. J., Hales, J., Jenkins, P., Keck, R., Ko, H., Molnar, G., Moss, E., Nielsen, P. P., Niendorf, K., Tombs, V., Warnick, M. & Wu, D. (2022). Odd, spoof perfect factorizations. Journal of Number Theory 234, 31–47. 10.1016/j.jnt.2021.07.028 · Checked 15 Oct 2026
- Ochem, P. & Zelinsky, J. (2026). On odd perfect numbers with exactly one even exponent greater than 2. arXiv:2607.19746 (v1, 22 July 2026). Preprint, not peer reviewed. Source · Checked 15 Oct 2026
- Acquaah, P. & Konyagin, S. (2012). On prime factors of odd perfect numbers. International Journal of Number Theory 8(6), 1537–1540. 10.1142/S1793042112500935 · Checked 15 Oct 2026
Source
Open, as listed by Wikipedia · Unsolved problems in mathematics. Checked 08 Oct 2026.
The status is the source’s. Papex does not decide whether a problem is solved.
Records on this problem
No public record addresses this problem yet.