Daily Silicon Valley

Daily Magazine For Entrepreneurs

Home » A Primer to Diophantine Analysis for Silicon Valley Scientists and Engineers

A Primer to Diophantine Analysis for Silicon Valley Scientists and Engineers

Author: Dr. Jonathan Kenigson, FRSA

Fermat’s Last Theorem is perhaps the most prominent example of a Diophantine equation, which is a polynomial equation with integer (whole number) coefficients that is to be solved over the integers and not a larger domain [1]. For instance, one may inquire which integers satisfy the equation x+2y=13 – a task that often conjures prime numbers, factorization, and abstract algebra [2]. The field is named in honor of the Greek-African mathematician Diophantus the Alexandrian, who lived in the 3rd century AD and wrote a series of books called Arithmetica [3]. During the Islamic Golden Age, mathematicians such as al-Khwarizmi, al-Karaji, and al-Haytham developed methods for solving linear and quadratic Diophantine equations and progressed towards solving higher degree equations [4]. In the 18th and 19th centuries, Euler [5], Lagrange [6], Legendre [7], and Gauss [8] studied Diophantine equations using modular arithmetic and the rudiments of the modern theory of algebraic factorization over finite fields. In the 20th century, Hilbert [9], Taniyama [10], and Shimura [10] advanced the field using algebraic geometry, modular forms, elliptic curves, and Galois theory. Roughly, Galois theory is the study of where certain polynomial functions “vanish” (or become zero) [11]. These points form algebraic structures in their own right, called Galois groups [12]. Also speaking crudely, elliptic curves are graphs of certain cubic equations (e.g. equations of degree 3) [13]. Andrew Wiles’s proof of Fermat’s Last Theorem relies heavily upon Taniyama and Shimura’s work. The Taniyama-Shimura conjecture states that every elliptic curve over the rational numbers (fractions as we typically use them) is a modular form [10]. Wiles’ proof of Taniyama-Shimura implies Fermat’s Last Theorem [14]. 

The first step in Wiles’ proof is to establish Taniyama-Shimura for a special class of elliptic curves that are called “semistable” [15]. Semistability is a condition that defines the curve’s behavior at prime numbers. When studying elliptic curves over number fields, it’s common to extend the base field to include all the algebraic numbers that are needed to define the curve [16]. In this setting, a prime p is said to be a prime of bad reduction if the reduction of the curve modulo p is not a smooth curve [17][18]. For example, if the coefficients have a common factor, then the equation defines a curve that is not smooth and therefore not an elliptic curve in the usual sense. A semistable elliptic curve is one that has good reduction at all but finitely many primes, and at the bad primes, the reduction is ‘as simple as possible’, meaning that the singularities are not mathematically intractable [17]. In particular, a semistable elliptic curve has only nodes or cusps as singularities at the bad primes [17]. Wiles’ proof technique exploits Galois representations and modular forms [10]. Fermat’s Last Theorem follows for all exponents except 4, a case that was later resolved by Wiles and Taylor [19]. By using modular forms in cryptography, Diophantine equations are used to create secure encryption schemes. In coding theory, they are used to design error-correcting codes [20]. Diophantine equations also reciprocally enrich group theory via the representations of groups of rational points on algebraic curves [21].

As an example, The Mordell-Weil Theorem describes the structure of the group of rational points on an elliptic curve [22]. Roughly, the Theorem states that the group of rational points on an elliptic curve over a number field (a finite extension of the rational numbers) is a finitely generated abelian group [22]. In other words, it says that the set of rational points on an elliptic curve is generated by a finite number of points. It allows one to determine whether a given Diophantine equation has a solution in rational numbers by studying the rational points on the associated elliptic curve [22]. The idea is to first find a set of generators for the group of rational points on the elliptic curve, and then use these generators to construct all rational solutions to the Diophantine equation [22]. Beside Mordell-Weil, another foundational result, Siegel’s Theorem, gives an upper bound on the number of integer solutions of a Diophantine equation in terms of the degree of the equation [23]. This trajectory of thought began in the early 20th century as Carl Ludwig Siegel concerned himself with bounding the number of integral (that is, whole-number) solutions to certain types of Diophantine equations that Gauss had discussed in terms of pure number theory [24]. Modern formulations of Seigel exploit the topological properties of the curve genus, a metric for the complexity of a curve. For an algebraic curve C of genus g > 0 defined over the field of rational numbers, there are only finitely many integral points on C. In other words, if the curve has a genus greater than 0, there are only a finite number of points on the curve with integer coordinates that lie on the graph [25]. The proof of Siegel’s Theorem relies on techniques from algebraic geometry, geometry of numbers, and the theory of Diophantine approximation [26]. The original proof by Siegel was quite involved, but later refinements and simplifications have been made by Faltings [27], Zannier [28], and Corvaja [29] since the 1980s.

Recall that algebraic numbers are the roots of non-zero polynomial equations with integer coefficients. Another result partially attributed to Siegel, The Thue-Siegel-Roth Theorem, gives absolute bounds on the quality of rational approximations to algebraic numbers [30]. The bounds furnished are not those imposed by limits of computational power. Rather, they are theoretical bounds imposed by the structure of the real numbers that cannot be transcended either by technological innovation or clever use of existing algorithms. Thue-Siegel-Roth states that algebraic numbers of degree greater than 1 cannot be approximated “too closely” by rational numbers. Diophantine equations help researchers understand the distribution and scarcity of these rational approximations to algebraic numbers [30]. It also has applications in other areas of mathematics, such as algebraic geometry, transcendental number theory, and the study of continued fractions [31]. Thue-Siegel-Roth Theorem is a strengthening of the earlier result known as Liouville’s Theorem [32], which states that any algebraic number α of degree d > 1 can only be approximated by rational numbers p/q to within |α – p/q| > C(α) / q^d, where C(α) depends upon α. Thue-Siegel-Roth improves this bound to q^(2-ε), which is significantly stronger for numbers of higher degree [30]. The Bombieri-Lang Conjecture, is a deep conjecture about the distribution of rational points on algebraic varieties that would furnish another general paradigm for looking at Diophantine equations [33]. Bombieri-Lang is an active area of research in computational and pure number theory to which Silicon Valley could contribute computational power and innovative ideas from computer science and simulation.


[1]. Lazebnik, Felix. “On systems of linear Diophantine equations.” Mathematics Magazine 69.4 (1996): 261-266.

[2]. Crescenzo, Peter. “A diophantine equation which arises in the theory of finite groups.” Advances in mathematics 17.1 (1975): 25-29.

[3]. Sesiano, Jacques. Books IV to VII of Diophantus’ Arithmetica: in the Arabic Translation Attributed to Qustā ibn Lūqā. Vol. 3. Springer Science & Business Media, 2012.

[4]. Christianidis, Jean, and Athanasia Megremi. “Tracing the early history of algebra: Testimonies on Diophantus in the Greek-speaking world (4th–7th century CE).” Historia Mathematica 47 (2019): 16-38.

[5]. Dujella, Andrej. “An extension of an old problem of Diophantus and Euler.” Fibonacci Quarterly 37 (1999): 312-314.

[6]. Mollin, R. A. “Generalized Lagrange criteria for certain quadratic Diophantine equations.” New York J. Math 11 (2005): 539-545.

[7]. Hata, Masayoshi. “Legendre type polynomials and irrationality measures.” (1990): 99-125.

[8]. Harman, Glyn. “Diophantine approximation with Gaussian primes.” The Quarterly Journal of Mathematics 70.4 (2019): 1505-1519.

[9]. Davis, Martin, Yuri Matijasevič, and Julia Robinson. “Hilbert’s tenth problem. Diophantine equations: positive aspects of a negative solution.” American Math. Soc Providence 1 (1976).

[10]. Darmon, Henri, and Claude Levesque. “Infinite Sums, Diophantine Equations and Fermat’s Last Theorem.” Number Theoretic Methods: Future Trends (2002): 73-95.

[11]. Morandi, Patrick. Field and Galois theory. Vol. 167. Springer Science & Business Media, 2012.

[12]. Stauduhar, Richard P. “The determination of Galois groups.” Mathematics of computation 27.124 (1973): 981-996.

[13]. Milne, James S. Elliptic curves. 2006.

[14]. Ribet, Kenneth A. “From the Taniyama-Shimura conjecture to Fermat’s last theorem.” Annales de la Faculté des sciences de Toulouse: Mathématiques. Vol. 11. No. 1. 1990.

[15]. Wiles, Andrew. “Modular elliptic curves and Fermat’s last theorem.” Annals of mathematics 141.3 (1995): 443-551.

[16]. Derickx, Maarten, et al. “Torsion points on elliptic curves over number fields of small degree.” arXiv preprint arXiv:1707.00364 (2017).

[17]. Koutsianas, Angelos. “Computing all elliptic curves over an arbitrary number field with prescribed primes of bad reduction.” Experimental Mathematics 28.1 (2019): 1-15.

[18]. Cremona, John E., and Mark P. Lingham. “Finding all elliptic curves with good reduction outside a given set of primes.” Experimental Mathematics 16.3 (2007): 303-312.

[19]. Calegari, Frank, and David Geraghty. “Modularity lifting beyond the Taylor–Wiles method.” Inventiones mathematicae 211.1 (2018): 297-433.

[20]. Kapoor, Vivek, Vivek Sonny Abraham, and Ramesh Singh. “Elliptic curve cryptography.” Ubiquity 2008.May (2008): 1-8.

[21]. Browning, T. D., D. R. Heath-Brown, and Per Salberger. “Counting rational points on algebraic varieties.” (2006): 545-578.

[22]. Brumer, Armand, and Oisín McGuinness. “The behavior of the Mordell-Weil group of elliptic curves.” (1990): 375-382.

[23]. Goldfeld, Dorian M. “A simple proof of Siegel’s theorem.” Proceedings of the National Academy of Sciences 71.4 (1974): 1055-1055.

[24]. Silverman, Joseph H. “A quantitative version of Siegel’s theorem: integral points on elliptic curves and Catalan curves.” (1987): 60-100.

[25]. Cappell, Sylvain E., and Julius L. Shaneson. “Genera of algebraic varieties and counting of lattice points.” Bulletin of the American Mathematical Society 30.1 (1994): 62-69.

[26]. Serre, Jean-Pierre, Martin Brown, and Michel Waldschmidt. Lectures on the Mordell-Weil theorem. F. Vieweg, 1989.

[27]. Faltings, Gerd, et al. Rational points. Vol. 6. Braunschweig‐Wiesbaden: Vieweg, 1986.

[28]. Corvaja, Pietro, and Umberto Zannier. “A subspace theorem approach to integral points on curves.” Comptes Rendus Mathematique 334.4 (2002): 267-271.

[29]. Corvaja, Pietro, and Umberto Zannier. “On the number of integral points on algebraic curves.” (2003): 27-42.

[30]. Chudnovsky, G. V. “The Thue-Siegel-Roth theorem for values of algebraic functions.” (1983): 281-284.

[31]. Töpfer, Thomas. “On the transcendence and algebraic independence of certain continued fractions.” Monatshefte für Mathematik 117 (1994): 255-262.

[32]. Nelson, Edward. “A proof of Liouville’s theorem.” Proc. AMS. Vol. 12. 1961.

[33]. Bresciani, Giulio. “On the Bombieri–Lang conjecture over finitely generated fields.” Algebra & Number Theory 16.10 (2023): 2409-2414.

Silicon Valley Daily

Daily magazine for entrepreneurs and business owners

Back to top