OpenFrontiers
← Back to the index
OP-12043OpenNumber theory

Ihara constant over F2\mathbb{F}_2 (C33C_{33})

Ihara constant over F2\mathbb{F}_2: determine C33C_{33} or improve its bounds; the repository records 0.316999... (lower) and 2−1≈0.41421\sqrt{2}-1 \approx 0.41421 (upper).

Editorial MSC2020 classification. Original article codes and source evidence are recorded in the literature directory.

Licensed source statementAs of Repository table at commit 2c1968c (2026-09-26); open while the recorded lower and upper bounds differSource & evidence ↓

Problem statement

C33=A(2)C_{33}=A(2) is the Ihara constant over F2\mathbb{F}_2. [DM2013-def-Aq]

For each integer g≥1g\ge 1, let

N2(g):=max⁡{#X(F2)  :  X/F2 a smooth projective geometrically integral curve of genus g}.N_{2}(g) := \max\bigl\{\#X(\mathbb{F}_2)\;:\; X/\mathbb{F}_2 \text{ a smooth projective geometrically integral curve of genus } g\bigr\}.

[DM2013-def-Nqg]

Then

A(2):=lim sup⁡g→∞N2(g)g.A(2) := \limsup_{g\to\infty}\frac{N_{2}(g)}{g}.

[DM2013-def-Aq]

Lean 4

Not yet attached

No Lean statement or proof is attached yet. Have a formalization? Share its source to add it to this problem.

Background

  • Status. The exact value of A(2)A(2) is unknown. As of the survey Bee2022, one has
0.316999… ≤ A(2) ≤ 2−1≈0.41421.0.316999\ldots \ \le\ A(2)\ \le\ \sqrt2-1\approx 0.41421.

[Bee2022-nonsquare-open] [DM2013-thm1.1] [DM2013-dv-bound]

  • Square vs. non-square fields. For qq a square, the Drinfeld--Vlăduţ upper bound is sharp: A(q)=q−1A(q)=\sqrt q-1, via explicit towers of function fields (e.g. the Garcia--Stichtenoth tower) GS1995. For non-square qq (in particular q=2q=2), no exact value is known. [DM2013-square-q] [Bee2022-gs-tower] [Bee2022-nonsquare-open]

  • Connection to coding theory (TVZ bound). The quantity A(q)A(q) controls the asymptotic performance of algebraic-geometry codes. In particular, the Tsfasman--Vlăduţ--Zink bound expresses an asymptotic rate--distance tradeoff in terms of A(q)A(q); see TVZ1982, Sti2009. [Sti2005-TVZ] [Bee2022-coding-theory]

  • Tables for fixed genus. For small genera, records and exact maxima for N2(g)N_2(g) are tabulated at manypoints.org. [Fab2022-manypoints]

  • Surveys and background: Bee2022, Sti2009.

References (from the repository page)

  • [Bee2022] Beelen, Peter. A survey on recursive towers and Ihara's constant. Preprint (2022). Google Scholar. arXiv PDF.

  • [Bee2022-lb-2-9] loc: arXiv v1 PDF p.3, Section 2.1 (Bounds using class field theory). quote: “Serre already demonstrated in his Harvard lectures, that for q=2q = 2, class field theory can be used to show that A(2)≥2/9A(2) \ge 2/9 [57, Theorem 5.11.1], a result that was also obtained using a different construction in [55].”

  • [Bee2022-lb-81-317] loc: arXiv v1 PDF p.3, Section 2.1 (Bounds using class field theory). quote: “The following table gives an overview: A(2)≥2/9≈0.222222…A(2) \ge 2/9 \approx 0.222222\ldots [57, 55] A(3)≥62/163≈0.380368…A(3) \ge 62/163 \approx 0.380368\ldots [53] A(2)≥81/317≈0.255520…A(2) \ge 81/317 \approx 0.255520\ldots [53] A(3)≥8/17≈0.470588…A(3) \ge 8/17 \approx 0.470588\ldots [59, 3]”

  • [Bee2022-lb-best] loc: arXiv v1 PDF p.3, Section 2.1 (Bounds using class field theory). quote: “The following table gives an overview: A(2)≥39/129≈0.302325…A(2) \ge 39/129 \approx 0.302325\ldots [45] A(3)≥0.492876…A(3) \ge 0.492876\ldots [25] A(2)≥0.316999…A(2) \ge 0.316999\ldots [25] The lower bounds for A(2)A(2) and A(3)A(3) found in [25] are currently the best known.”

  • [Bee2022-gs-tower] loc: arXiv v1 PDF p.6, Section 3.1 (The first two Garcia–Stichtenoth towers). quote: “As n→∞n \to \infty the ratio of number of places of degree one of EnE_n, denoted by N1(En)N_1(E_n), and the genus of EnE_n, denoted by g(En)g(E_n), tends to q−1q - 1, achieving the Drinfeld–Vladut bound.”

  • [Bee2022-nonsquare-open] loc: arXiv v1 PDF p.19, Section 5 (Recursive towers of function fields: non-square finite fields). quote: “The case of non-square finite fields is currently still open.”

  • [Bee2022-coding-theory] loc: arXiv v1 PDF p.4, Section 2.2 (Explicit equations for modular curves). quote: “One of the driving motivations for studying families of curves with many Fq\mathbb{F}_q-rational points, is that using Goppa's construction of error-correcting codes, such families can be used to find good families of such codes [62].”

  • [DM2013] Duursma, Iwan; Mak, Koon-Ho. On lower bounds for the Ihara constants A(2)A(2) and A(3)A(3). Compositio Mathematica 149 (2013), 1108–1128. DOI: 10.1112/S0010437X12000796. Google Scholar. arXiv PDF

  • [DM2013-weil-bound] loc: arXiv v4 PDF p.1, Section 1 (Introduction). quote: “It is well-known that the Weil bound #X(Fq)≤q+1+2gq\# X(\mathbb{F}_q) \le q + 1 + 2g\sqrt{q} is not sharp if gg is large compared to qq.”

  • [DM2013-def-Nqg] loc: arXiv v4 PDF p.1, Section 1 (Introduction). quote: “Put Nq(g):=max⁡#X(Fq)N_q(g) := \max \#X(\mathbb{F}_q), where the maximum is taken over all curves X/FqX/\mathbb{F}_q with genus gg.”

  • [DM2013-def-Aq] loc: arXiv v4 PDF p.1, Section 1 (Introduction). quote: “The Ihara constant is defined by A(q):=lim sup⁡g→∞Nq(g)/gA(q) := \limsup_{g\to\infty} N_q(g)/g.”

  • [DM2013-dv-bound] loc: arXiv v4 PDF p.1, Section 1 (Introduction). quote: “For any qq, we have A(q)≤q−1A(q) \le \sqrt{q} - 1 (see [4]).”

  • [DM2013-square-q] loc: arXiv v4 PDF p.1, Section 1 (Introduction). quote: “and if qq is a square we have (see [12, 28]) A(q)=q−1A(q) = \sqrt{q} - 1.”

  • [DM2013-prop1.1] loc: arXiv v4 PDF p.2, Section 1 (Introduction). quote: “Among these results, the best lower bounds are A(2)≥97/376=0.257979…A(2) \ge 97/376 = 0.257979\ldots by Xing and Yeo [31], and A(3)≥12/25=0.48A(3) \ge 12/25 = 0.48 by Atiken and Hajir [8].”

  • [DM2013-prop1.2] loc: arXiv v4 PDF p.2, Section 1 (Introduction). quote: “In [13], Kuhnt obtained a better lower bound for A(2)A(2), which says A(2)≥39/129=0.302325…A(2) \ge 39/129 = 0.302325\ldots.”

  • [DM2013-thm1.1] loc: arXiv v4 PDF p.2, Section 1 (Introduction). quote: “Theorem 1.3. A(2)≥0.316999…A(2) \ge 0.316999\ldots.”

  • [DV1983] Drinfeld, V. G.; Vladut, S. G. Number of points of an algebraic curve. Functional Analysis and Its Applications 17 (1983), no. 1, 53–54. DOI: 10.1007/BF01083182. Google Scholar.

  • [Fab2022] Faber, Xander; Grantham, Jon. Binary curves of small fixed genus and gonality with many rational points. Journal of Algebra 597 (2022), 24–46. DOI: 10.1016/j.jalgebra.2022.01.008. Google Scholar. arXiv PDF

  • [Fab2022-manypoints] loc: arXiv v2 PDF p.1, Section 1 (Introduction). quote: “Write N2(g)N_2(g) for the maximum number of rational points on a curve of genus gg over F2\mathbb{F}_2. The values in Table 1 come from manypoints.org.”

  • [GS1995] Garcia, Arnaldo; Stichtenoth, Henning. A tower of Artin–Schreier extensions of function fields attaining the Drinfeld–Vladut bound. Inventiones Mathematicae 121 (1995), 211–222. DOI: 10.1007/BF01884295. Google Scholar.

  • [NX1998] Niederreiter, Harald; Xing, Chaoping. Towers of global function fields with asymptotically many rational places and an improvement of the Gilbert–Varshamov bound. Mathematische Nachrichten 195 (1998), 171–186. DOI: 10.1002/mana.19981950110. Google Scholar.

  • [Sch1992] Schoof, Rene. Algebraic curves over F2\mathbb{F}_2 with many rational points. Journal of Number Theory 41 (1992), no. 1, 6–14. DOI: 10.1016/0022-314X(92)90079-5. Google Scholar. Author PDF

  • [Ser1983] Serre, Jean-Pierre. Sur le nombre des points rationnels d'une courbe algebrique sur un corps fini. C. R. Acad. Sci. Paris Ser. I Math. 296 (1983), 397–402. Google Scholar. Publisher entry

  • [Sti2005] Stichtenoth, Henning. Transitive and self-dual codes attaining the Tsfasman-Vladut-Zink bound. IEEE Transactions on Information Theory 52 (5) (2006), 2218–2224. DOI: 10.1109/TIT.2006.872986. Google Scholar. arXiv PDF

  • [Sti2005-TVZ] loc: arXiv v1 PDF p.2, Section 1 (Introduction and Main Results). quote: “Then αq(δ)≥1−δ−A(q)−1\alpha_q(\delta) \ge 1 - \delta - A(q)^{-1} for 0≤δ≤10 \le \delta \le 1.”

  • [Sti2009] Stichtenoth, Henning. Algebraic Function Fields and Codes. 2nd ed., Graduate Texts in Mathematics 254, Springer (2009). DOI: 10.1007/978-3-540-76878-4. Google Scholar.

  • [TVZ1982] Tsfasman, M. A.; Vladut, S. G.; Zink, T. Modular curves, Shimura curves, and Goppa codes, better than the Varshamov–Gilbert bound. Mathematische Nachrichten 109 (1982), 21–28. DOI: 10.1002/mana.19821090103. Google Scholar.

  • [XY2007] Xing, Chaoping; Yeo, Sze Ling. Algebraic curves with many points over the binary field. Journal of Algebra 311 (2007), no. 2, 775–780. DOI: 10.1016/j.jalgebra.2006.12.029. Google Scholar.

Progress & source assessment

Repository table (commit 2c1968c, 26 September 2026): best known lower bound 0.316999...; best known upper bound 2−1≈0.41421\sqrt{2}-1 \approx 0.41421. Values marked with an asterisk have minimal verification.

Known upper bounds

Bound Reference Comments
22≈2.828432\sqrt2 \approx 2.82843 Classical (Weil bound) From #X(F2)≤2+1+2g2\#X(\mathbb{F}_2)\le 2+1+2g\sqrt2, hence N2(g)g≤22+3g\frac{N_2(g)}{g}\le 2\sqrt2+\frac{3}{g}. [DM2013-weil-bound]
2−1≈0.41421\sqrt2-1 \approx 0.41421 DV1983 Drinfeld--Vlăduţ (Ihara) bound: A(q)≤q−1A(q)\le \sqrt q-1 for every prime power qq. [DM2013-dv-bound]

Known lower bounds

Bound Reference Comments
00 Trivial Since N2(g)≥0N_2(g)\ge 0.
2/9≈0.222222/9 \approx 0.22222 Ser1983, Sch1992 Serre's class field tower method; explicit constructions over F2\mathbb{F}_2. [Bee2022-lb-2-9]
81/317≈0.2555281/317 \approx 0.25552 NX1998 Improves the 2/92/9 bound. [Bee2022-lb-81-317]
97/376≈0.2579897/376 \approx 0.25798 XY2007 Improves the 81/31781/317 bound. [DM2013-prop1.1]
39/129≈0.3023339/129 \approx 0.30233 DM2013 Lower bound reported in DM2013 (attributed there to Kuhnt's thesis). [DM2013-prop1.2]
0.316999…0.316999\ldots DM2013 Lower bound from DM2013 (also listed in Bee2022). [DM2013-thm1.1] [Bee2022-lb-best]
References
  1. [1]Davis, Damek; Ivanisvili, Paata; Tao, Terence (maintainers). Optimization Constants in Mathematics, github.com/teorth/optimizationproblems (commit 2c1968c, 26 September 2026). — Ihara constant over $\mathbb{F}_2$ ($C_{33}$) linkApache-2.0 (repository license); constant pages reproduced with attribution; repository references retained.

Recorded progress

Add yours →

Nothing recorded yet. A pointer to a paper the entry is missing counts.

Machine attempts

0 recorded

Attempts run against the statement below and are recorded in full, including the failures. The point is to stop the next model — and the next person — from re-walking a road that is already known to end. Each attempt costs 25 credits.

No machine attempts recorded.

Sign in to run an attempt. New accounts start with enough credits for ten.

Discussion

0 notes

Notes are for clarifying the statement, flagging a missing reference, or recording why an approach fails. Claimed solutions belong in a submission, where they can be reviewed.

No notes yet.

Sign in to add a note.