Proth prime
A Proth prime is a prime number of the form k \cdot 2^n + 1, where k is an odd positive integer, n is a positive integer, and k < 2^n.[1] These primes are named after the French mathematician François Proth, who introduced the form in his 1878 paper "Théorèmes sur les nombres premiers," published in the Comptes rendus hebdomadaires des séances de l'Académie des sciences.[2] In that work, Proth also established a key result now known as Proth's theorem, which provides a deterministic criterion for verifying the primality of such numbers.[3] Specifically, for a Proth number N = k \cdot 2^n + 1, if there exists an integer a coprime to N such that a^{(N-1)/2} \equiv -1 \pmod{N}, then N is prime.[1] This condition leverages properties of quadratic residues and Euler's criterion, making it an efficient test for numbers in this form, particularly when n is large.[4] Proth primes encompass several notable subclasses, including Fermat primes (where k = 1 and n = 2^m for some integer m) and certain Cullen primes (where k = n).[2] The smallest Proth primes are 3, 5, 13, 17, 41, and 97.[5] Their structure facilitates computational primality proving, which has made them central to distributed computing efforts in prime hunting, such as PrimeGrid's Proth Prime Search subproject, which continues to seek large instances as of 2025. The largest known Proth prime is $10223 \cdot 2^{31172165} + 1, with 9,383,761 digits, discovered by PrimeGrid in 2016 and remaining the record as of November 2025.[6][7] Despite their utility, fundamental questions remain unresolved, including whether infinitely many Proth primes exist—a conjecture tied to broader problems in analytic number theory, such as the distribution of primes in arithmetic progressions.[2] Proth primes also play a role in studying Sierpinski numbers, which are odd integers k for which k \cdot 2^n + 1 is composite for all n \geq 1; finding Proth primes helps bound or resolve candidates in these searches.[4]Definition and Background
Definition
A Proth prime is a prime number of the form N = k \cdot 2^n + 1, where k is an odd positive integer, n is a positive integer, and $2^n > k.[5] This specific form distinguishes Proth primes as a subset of primes that exhibit a structured binary representation, with the majority of bits set to zero except for a small odd multiplier k followed by n zeros in binary and ending with a 1.[2] The parameter k is required to be odd to ensure the number aligns with the conventional definition of Proth forms, avoiding even multipliers that could lead to alternative factorizations or overlap with other prime classes.[5] The condition $2^n > k guarantees that the exponent n dominates the structure, preventing the form from reducing to simpler cases like small primes or Fermat numbers without the full Proth characterization.[8] Meanwhile, n as a positive integer allows inclusion of foundational examples while maintaining the form's generality. The smallest Proth prime is 3, obtained with k = 1 and n = 1 since $1 \cdot 2^1 + 1 = 3 and $2^1 > 1.[5] Another early example is 5, with k = 1 and n = 2 ($1 \cdot 2^2 + 1 = 5, $2^2 > 1). These boundary cases illustrate how the form encompasses small primes while scaling to larger ones, such as 13 (k = 3, n = 2). Proth primes are named after the French mathematician François Proth, who first studied numbers of this form and developed a primality test for them in his 1878 paper "Théorèmes sur les nombres premiers."[2]Historical Context
François Proth (1852–1879) was a self-taught French mathematician and farmer residing in Vaux-devant-Damloup near Verdun, who contributed to number theory without formal academic training. Despite his brief life, ending at age 26, Proth published several theorems on primes, drawing from contemporary studies on arithmetic progressions and special forms. His most notable work appeared in 1878, when he introduced a class of numbers of the form k \cdot 2^n + 1 (with k odd and n a positive integer) and devised a deterministic primality test applicable under certain conditions, marking an early insight into efficient verification for such candidates.[9] Proth's contributions received limited immediate attention amid 19th-century explorations of prime distributions, such as those by Chebyshev and others on forms like k \cdot 2^n \pm 1. Subsequent extensions emerged sporadically: in 1886, Paul Seelhoff compiled a list of 25 primes of this form, including examples up to $5 \cdot 2^{39} + 1, often in connection with Fermat number factorizations. By 1904, H. J. A. Cunningham and A. E. Western had identified 720 such primes, expanding tables to support factorization efforts, while Maurice Kraitchik discovered in 1925 that the Proth prime $579 \cdot 2^{21} + 1 divides the Fermat number F_{15}. These efforts highlighted ties to broader prime form investigations but did not elevate Proth's test to prominence.[10] The significance of Proth primes waned until the mid-20th century, when computational number theory revived interest in the 1970s and 1980s amid growing capabilities for large-integer arithmetic. Projects like those by Will Keller extended search tables (e.g., for k < 20 and n \leq 8500 by 1983), revealing new primes and factors of Fermat numbers, thus integrating Proth's method into systematic hunts. The question of whether infinitely many Proth primes exist, unresolved since the era's prime form studies, persists as an open problem in analytic number theory. In the 1990s, milestones included Yves Gallot's development of the proth.exe program, which leveraged Proth's theorem for efficient testing and verified numerous large instances, such as those exceeding thousands of digits, accelerating discoveries in distributed computing efforts.[10][2][11]Properties and Relations
Proth Numbers
Proth numbers are natural numbers of the form N = k \times 2^n + 1, where k is an odd positive integer, n is a positive integer with n > 0, and $2^n > k. This general class includes both primes and composites meeting these criteria, with Proth primes comprising the prime instances.[12] Proth numbers are always odd, as k \times 2^n is the product of an odd integer and an even power of 2 (for n \geq 1), yielding an even result, and adding 1 produces an odd integer. When the exponent n is composite, Proth numbers can be factored algebraically under certain conditions, such as aurifeuillian factorizations applicable to specific values of k.[13] For instance, 9 = $1 \times 2^3 + 1 is a composite Proth number. The distribution of Proth numbers is sparser than that of ordinary primes owing to the constraints of their form. The sum of the reciprocals of Proth primes converges to a value between 0.7473924793 and 0.7473924795, as determined in 2022, and no proof exists for the infinitude of Proth primes.[2]Connections to Other Prime Forms
Proth primes exhibit notable connections to Fermat primes, as the latter form a specific subset of the Proth form. Fermat primes, defined as numbers of the form $2^{2^m} + 1 for nonnegative integers m, correspond to Proth primes with k = 1 and exponent n = 2^m, satisfying the condition $2^n > k.[14] This inclusion highlights how Fermat primes represent the canonical case where the odd multiplier is unity, and their rarity—only five known examples—underscores the sparsity within the broader Proth class.[5] A deeper relation arises through the factorization of Fermat numbers. Euler's theorem states that any prime divisor p of the Fermat number F_m = 2^{2^m} + 1 (for m > 0) must be of the form p = k \cdot 2^{m+2} + 1, where k is a positive odd integer.[14] Such divisors are thus Proth primes provided k < 2^{m+2}, a condition often met for the algebraic factors discovered in composite Fermat numbers. For instance, the factor 579 · 2^{21} + 1 of F_{15} exemplifies this, as identified by Kraitchik in 1926, demonstrating how searches for Proth primes contribute to Fermat number factorizations.[10] Similarly, 5 · 2^{39} + 1 divides F_{36}, found by Seelhoff in 1886.[10] Proth primes also intersect with Cullen and Sierpiński primes through shared structural forms. Cullen primes, given by n \cdot 2^n + 1 for positive integers n > 1, constitute a subset of Proth primes, where k = n and the exponent equals k, adhering to $2^n > n.[14] This overlap implies that discoveries of Cullen primes simultaneously yield Proth primes, aiding searches in both categories. In contrast, Sierpiński primes relate inversely: a Sierpiński number k is an odd positive integer such that k \cdot 2^n + 1 is composite for all n \geq 1, often proven via covering systems of congruences.[14] The existence of a Proth prime for a candidate k (i.e., k \cdot 2^n + 1 prime for some n) eliminates that k from being Sierpiński, as in the Seventeen or Bust project targeting remaining candidates.[14] Thus, Proth prime hunts directly probe the Sierpiński conjecture's boundaries.[15] Extending to generalized Fermat primes, these are numbers of the form k \cdot 2^{2^m} + 1 with odd k > 1, generalizing the Fermat case. Proth primes encompass this form when the exponent $2^m satisfies the size condition relative to k, positioning Proth primes as a foundational class for such generalizations.[5] Factors of generalized Fermat numbers a^{2^m} + 1 (for odd a > 1) are likewise Proth primes, as established by Riesel in 1969, with examples including divisors found by Dubner and Keller in 1995 for a = 6, 10, 12.[10] This connection extends to Cunningham chains, sequences of primes where each subsequent member is $2p_i + 1 (first kind) or $2p_i - 1 (second kind); Proth primes frequently appear in such chains due to their algebraic structure, facilitating extensions in prime tuple searches.[10] Theoretically, Proth primes play a role in covering systems and conjectures on prime distributions. In the Sierpiński problem, covering systems—sets of moduli that congruentially cover all integers n—render k \cdot 2^n + 1 divisible by small primes, producing composites; Proth primes evade such coverings for specific k, testing the conjecture's scope.[15] Regarding prime gaps, while not central, the form's sparsity informs bounds on gaps in thin sets, as Proth primes contribute to understanding irregularities in prime occurrences.[5] More broadly, the infinitude of Proth primes remains open but is supported by heuristic densities akin to Dirichlet's theorem on primes in arithmetic progressions; explicit bounds on the reciprocal sum \sum_{p \in \mathcal{P}} \frac{1}{p} (where \mathcal{P} is the set of Proth primes) between 0.7473924793 and 0.7473924795, as determined in 2022, indicate sparsity consistent with conjectured infinitude.[16]Primality Testing
Proth's Theorem
Proth's theorem, proposed by François Proth in 1878, provides a sufficient criterion for determining the primality of numbers of the form p = k \times 2^n + 1, where k is a positive odd integer, n \geq 1, and k < 2^n. The theorem asserts that p is prime if there exists an integer a such that a^{(p-1)/2} \equiv -1 \pmod{p}. This condition leverages Euler's criterion, which links the Legendre symbol \left( \frac{a}{p} \right) to the modular exponentiation for prime moduli.[17][2] The requirement that a be a quadratic non-residue modulo p is central, as Euler's criterion equates a^{(p-1)/2} \equiv \left( \frac{a}{p} \right) \pmod{p} when p is prime, yielding -1 precisely for non-residues. In practice, identifying such an a involves testing small candidates, often primes up to a modest bound like 29, since approximately half of the integers modulo p are non-residues. If the congruence holds for any such a, primality is confirmed; otherwise, further witnesses or alternative tests may be needed.[18][17] The proof hinges on properties of quadratic residues and the cyclic structure of the multiplicative group (\mathbb{Z}/p\mathbb{Z})^\times, which has order p-1 if p is prime. The congruence implies that the order of a divides p-1 but not (p-1)/2 = k \times 2^{n-1}, forcing the order to be exactly p-1, as the 2-primary part of p-1 is $2^n, exceeding the square root of p under the given constraints. This order condition cannot hold if p were composite, since the group order would then be strictly less than p-1. Modern expositions often frame this via Pocklington's theorem, but Proth's original argument directly exploits the form's algebraic structure.[17][19] While deterministic—requiring exhaustive search over possible a in the worst case—the theorem's efficiency diminishes for large n, as modular exponentiation scales poorly without fast algorithms like binary exponentiation. The criterion is sufficient but not necessary; its failure for tested bases neither confirms nor refutes primality, necessitating complementary methods for full verification.[17][18]Modern Algorithms
In contemporary primality testing for Proth numbers of the form N = k \cdot 2^e + 1 (with k odd and k < 2^e), deterministic algorithms extend the classical Proth test by leveraging algebraic structures for unconditional proofs. A key advancement is the 2008 algorithm by Tsz-Wo Sze, which deterministically verifies primality without unproven conjectures by computing square roots modulo N through group isomorphism and randomized integer selection to find a quadratic non-residue. For cases where k = O(\log N), the worst-case time complexity is \tilde{O}((\log N)^3) bit operations, assuming fast multiplication; broader cases yield \tilde{O}((k \log k + \log N) \log^2 N).[20] Probabilistic methods complement deterministic ones by providing rapid initial screening, particularly for large exponents where full proofs are resource-intensive. The Miller-Rabin test is frequently adapted for Proth candidates as a compositeness detector, performing O(r (\log N)^3) operations for r witnesses, with error probability below $4^{-r}; it exploits the special form to select witnesses efficiently. Lucas pseudoprime tests are integrated when the complete factorization of N-1 = k \cdot 2^e is available (trivial for small k), verifying conditions like U_{N-1}(P, Q) \equiv 0 \pmod{N} in O((\log N)^2) time per iteration. Preprocessing via the P-1 factoring algorithm can identify prime factors p of N for which p-1 is smooth with respect to chosen bounds, often eliminating composites before full testing. These algorithms achieve practical efficiency on modern hardware, where testing Proth numbers with exponents up to millions completes in hours for modest k (e.g., k < 1000) using optimized FFT-based arithmetic libraries. Parallelization via distributed computing distributes sieve-filtered candidates across volunteer networks, scaling searches to exponents exceeding 30 million through task partitioning.[21] Implementations are available in specialized software: Prime95 supports probable prime (PRP) testing with Miller-Rabin for Proth forms via its general-purpose engine; YAFU combines ECM/P-1 factoring with BPSW (Baillie-PSW) probabilistic verification, including Proth adaptations; and LLR (Lucas-Lehmer-Riesel) optimizes deterministic Proth tests for distributed projects. Custom scripts, often in C/GMP, enable tailored searches with integrated sieving.[21]Variants
Proth Primes of the Second Kind
Proth primes of the second kind are prime numbers of the form N = k \times 2^n - 1, where k is an odd positive integer, n > 1 is an integer, and $2^n > k. This form contrasts with the standard Proth primes, which add 1 rather than subtract it. These primes belong to a broader class analogous to Proth numbers but adapted for the minus sign, often studied in connection with sequences where primality depends on specific choices of k and n. A key property is that when k = 1, these reduce to Mersenne primes of the form $2^n - 1, for which n must be prime to yield a prime N. For general odd k > 1, however, n need not be prime; primes can occur even when n is composite, though algebraic factorizations are possible in such cases using identities for expressions like x^d - 1 where d divides n, potentially revealing non-trivial factors. For instance, if n = ab with a > 1, the number may factor via cyclotomic polynomials or aurifeuillian identities adapted to the form.[22] Primality testing for these numbers employs the Lucas-Lehmer-Riesel (LLR) test, analogous to Proth's theorem but tailored to the -1 case using Lucas sequences with parameters P and Q = -1. Select an odd integer P such that the discriminant D = P^2 + 4 is a quadratic non-residue modulo N. The Lucas V-sequence is defined by V_0 = 2, V_1 = P, and V_i = P V_{i-1} + V_{i-2} for i \geq 2. The test initializes s_0 as the appropriate term V_d(P, -1) \mod N (where d is the "setup exponent," often small and chosen to satisfy conditions), then iterates s_i = s_{i-1}^2 - 2 \mod N for i = 1 to n-2, checking if s_{n-2} \equiv 0 \pmod{N}. If yes, and N passes trial division, N is prime.[23] In contexts like the Riesel problem—seeking odd k where all k \times 2^n - 1 are composite for n \geq 1—a covering set of moduli is required to ensure every n falls into a residue class divisible by a fixed prime, proving compositeness without individual tests. As of November 2025, the smallest conjectured Riesel number is 509203, with ongoing searches by projects like PrimeGrid.[24] Representative small examples include $7 = 1 \times 2^3 - 1 and $11 = 3 \times 2^2 - 1, both prime. Larger instances appear in sequences for fixed small odd k, known collectively as Riesel primes, such as $23 = 3 \times 2^3 - 1. These examples illustrate how the form generates primes across varying k and n, with extensive searches revealing thousands up to significant sizes.Related Forms
Generalized forms of Proth primes extend the original structure k \times 2^n + 1 (with k odd and n > \log_2 k) to expressions like k \times b^n \pm 1, where b > 2 is an integer base, k is a positive integer, and n satisfies appropriate conditions for the form to be candidate for primality.[25] These generalizations allow primality testing via extensions of Proth's theorem, often employing Lucas-Lehmer-like sequences adapted to the base b.[25] When k = 1, these reduce to Cunningham numbers of the first or second kind, b^n + 1 or b^n - 1, which form chains in the Cunningham project for factoring and prime hunting.[26] Hybrid forms arise when parameters align with other named prime classes; for instance, Cullen primes n \times 2^n + 1 (with n odd) coincide with Proth primes under the condition $2^n > n.[14] Similarly, adjusted Woodall primes n \times 2^n - 1 (with n odd) overlap with Proth-like forms of the second kind, k \times 2^n - 1.[14] Theoretical extensions of Proth forms appear in algebraic contexts, such as generalizations using Gauss sums over finite fields to test primality for numbers of the form k \times p^n + 1, where p is an odd prime.[27] Further generalizations consider irreducibility and primality in rings of algebraic integers, tying Proth-like expressions to units in cyclotomic or quadratic fields, though explicit prime generation remains limited.[27] For higher-degree polynomials, extensions explore primes generated by trinomials or other forms analogous to k b^n \pm 1, but these lack dedicated testing theorems comparable to Proth's.[25] The infinitude of primes in these generalized Proth forms remains an open problem, mirroring the unresolved question for standard Proth primes, with heuristic evidence from reciprocal sums suggesting positive density but no proof.[2] Analogous conjectures apply to variants like k b^n - 1 (second kind) and higher-base forms, where Dirichlet's theorem provides infinitude in related arithmetic progressions but not for the exponential growth.[2]Applications
Prime Generation and Searches
The search for Proth primes typically involves fixing a small odd integer k and systematically increasing the exponent n to test candidates of the form k \cdot 2^n + 1, where $2^n > k, ensuring the numbers satisfy the Proth number condition. This strategy leverages the structure of Proth numbers for efficient sieving to eliminate composites early, followed by probable prime (PRP) testing on survivors. Double-checking of searched ranges is performed using advanced sieving techniques to confirm no primes were overlooked, often covering specific k and n intervals with high confidence.[28][29] Distributed computing projects have played a central role in these efforts since 1998, when the Proth Search project was established by Ray Ballinger and Wilfrid Keller to coordinate global searches for Proth primes. Complementing this, PrimeGrid's Proth Prime Search subproject, launched as part of its BOINC-based distributed computing framework, focuses on k < 300 and n up to several million, harnessing volunteer resources worldwide to test vast ranges. These initiatives employ tools like the Lucas-Lehmer-Riesel (LLR) test for PRP verification, enabling the processing of millions of candidates efficiently.[28][21][21] Key milestones include the completion of searches up to n \leq 7,000,000 for all k < 130 in July 2025, as coordinated through the Proth Search project. Earlier achievements encompass full coverage up to n \leq 6,500,000 for k < 166 by April 2025 and progressive extensions for larger k ranges via PrimeGrid efforts. Proth primes have also contributed to broader number theory verifications, such as the 2013 numerical confirmation of the ternary (weak) Goldbach conjecture up to approximately $8.875 \times 10^{30}, where prime ladders incorporating Proth primes facilitated the required prime chain constructions.[29][30] Challenges in Proth prime searches intensify with increasing size, as the density of composites rises, demanding extensive computational resources for sieving and testing large n. Typically, candidates are first identified as probable primes via PRP methods, but definitive certification requires additional rigorous primality proofs, such as those based on Proth's theorem, to confirm primality amid the growing scarcity of primes.[21][28]Cryptographic and Computational Uses
Proth primes find applications in cryptography due to their structural properties that enable efficient computations in finite fields. Their form k \cdot 2^n + 1, with k odd and k < 2^n, facilitates optimized modular reduction algorithms without pre-computation, which is particularly useful in elliptic curve cryptography (ECC) implementations over special prime fields, as discussed in contexts like FIPS 186-3, where techniques for forms including Proth primes reduce the overhead in scalar multiplications and point operations on elliptic curves.[31] This efficiency arises from the ability to perform Montgomery multiplication tailored to Proth moduli. Additionally, Proth primes optimize the denBoer reduction, which equates the computational complexity of the Diffie-Hellman problem to the discrete logarithm problem in certain elliptic curve groups; for instance, curves with Proth prime order points, like n = 1 + 55 \times 2^{286}, require fewer oracle calls (as low as 286) to solve the discrete logarithm using Diffie-Hellman queries, enhancing security analysis and protocol design.[32] In computational number theory, Proth primes and Proth's theorem play a key role in large-scale verifications of conjectures, exemplified by their use in confirming the weak Goldbach conjecture—that every odd integer greater than 5 is the sum of three primes. In 2013, numerical verification extended this up to $8.875 \times 10^{30} using a "prime ladder" algorithm that constructs sequences of primes with controlled gaps, building on prior checks of the binary Goldbach conjecture up to $4 \times 10^{18}; Proth's theorem enabled rapid primality testing for candidates of the form k \cdot 2^n + 1 with n \geq 52, sieving small divisors and leveraging the theorem's deterministic condition a^{(N-1)/2} \equiv -1 \pmod{N} where (a/N) = -1, far outperforming general tests like ECPP for these special forms.[30] Beyond these, Proth primes support efficient number theoretic transforms (NTTs) in lattice-based cryptosystems for secure signal processing, where moduli of the form q = k \cdot 2^l + 1 (with q \equiv 1 \pmod{2^n}) allow exact cyclic convolutions for polynomial multiplications in encrypted domains without rounding errors, as implemented in rings \mathbb{Z}_q / (x^n + 1). For example, the NIST-standardized CRYSTALS-Kyber (FIPS 203, 2024) employs the Proth prime q = 3329 = 13 \cdot 2^8 + 1 for its NTT in polynomial ring arithmetic.[33] They also aid in testing conjectures on prime densities, such as bounds on the reciprocal sum of Proth primes, providing empirical data on their distribution among numbers of this form.[2] The advantages of Proth primes stem from their ease of candidate generation—selecting small odd k and testing large n via Proth's theorem, which is computationally lightweight—and their algebraic structure, which supports fast modular arithmetic operations like exponentiation and inversion due to the sparse binary representation and smooth N-1 = k \cdot 2^n.[31]Known Examples and Records
Small Proth Primes
The smallest Proth primes provide foundational examples of numbers of the form k \cdot 2^n + 1, where k is an odd positive integer less than $2^n and n \geq 1, demonstrating the structure's potential to yield primes among otherwise sparse forms. These early instances, all below 2000, were identified as primes through simple verification methods available in the 19th century or earlier, predating François Proth's 1878 theorem on their primality testing.[5][10] The sequence begins with well-known small primes that fit the Proth form, such as 3 (k=1, n=1), 5 (k=1, n=2), and 13 (k=3, n=2). Subsequent examples include 17 (k=1, n=4) and 41 (k=5, n=3), which were recognized as primes via trial division by ancient and medieval mathematicians, though their classification as Proth numbers came later with the form's formal study. By the early 20th century, systematic manual computations by researchers like A. E. Western cataloged hundreds of such primes up to moderate sizes, confirming additional small ones like 97 (k=3, n=5) and 113 (k=7, n=4) through exhaustive factorization checks.[5][10] The first 20 Proth primes, as cataloged in the On-Line Encyclopedia of Integer Sequences (OEIS A080076), are listed below with their parameters and digit counts. These values highlight how small odd k values, particularly 1, 3, and 5, generate multiple early primes, reflecting a pattern of relative density at low scales that diminishes rapidly as n increases due to the exponential growth of the numbers and the inherent rarity of primes.[34][5]| k | n | Prime p | Digits |
|---|---|---|---|
| 1 | 1 | 3 | 1 |
| 1 | 2 | 5 | 1 |
| 3 | 2 | 13 | 2 |
| 1 | 4 | 17 | 2 |
| 5 | 3 | 41 | 2 |
| 3 | 5 | 97 | 2 |
| 7 | 4 | 113 | 3 |
| 3 | 6 | 193 | 3 |
| 15 | 4 | 241 | 3 |
| 1 | 8 | 257 | 3 |
| 11 | 5 | 353 | 3 |
| 7 | 6 | 449 | 3 |
| 9 | 6 | 577 | 3 |
| 5 | 7 | 641 | 3 |
| 21 | 5 | 673 | 3 |
| 3 | 8 | 769 | 3 |
| 29 | 5 | 929 | 3 |
| 9 | 7 | 1153 | 4 |
| 19 | 6 | 1217 | 4 |
| 11 | 7 | 1409 | 4 |
Largest Known Proth Primes
The largest known Proth prime remains $10223 \times 2^{31172165} + 1, with 9,383,761 decimal digits, discovered on October 31, 2016, by Szabolcs Péter through PrimeGrid's Seventeen or Bust subproject.[7][6] This discovery utilized distributed computing across volunteer GPUs for the initial search, followed by verification via the Lucas-Lehmer-Riesel (LLR) test, which required several weeks on multi-core systems to confirm primality.[7] At the time, it ranked as the seventh-largest known prime overall and the largest non-Mersenne prime.[6] Subsequent records have advanced the scale, though none have surpassed the 2016 milestone by November 2025. The second-largest is $81 \times 2^{20498148} + 1, featuring 6,170,560 digits and identified on June 13, 2023, by Ryan Propper using LLR software; this prime divides the generalized Fermat number F_2(3 \times 2^{5124537}).[35] The third-largest, $7 \times 2^{20267500} + 1 with 6,101,127 digits, was found on July 21, 2022, also by Propper via LLR and divides the Fermat number F_{20267499}(12).[36] These discoveries highlight the efficiency of LLR for Proth forms, often completed on high-end desktop hardware after targeted searches.[36][35] Distributed projects like PrimeGrid have driven progress, enabling exhaustive searches over large ranges of k and n through volunteer contributions.[21] From the early 2000s, when Proth primes around 1 million digits were records, sizes have expanded to over 9 million digits by leveraging GPU acceleration for sieving and testing, with formal proofs ensuring reliability.[21][37]| Rank | k | n | Digits | Date | Discoverer | Project/Source |
|---|---|---|---|---|---|---|
| 1 | 10223 | 31172165 | 9,383,761 | 2016-10-31 | Szabolcs Péter | PrimeGrid |
| 2 | 81 | 20498148 | 6,170,560 | 2023-06-13 | Ryan Propper | LLR |
| 3 | 7 | 20267500 | 6,101,127 | 2022-07-21 | Ryan Propper | LLR |