🔢 mathematics

Introduction to Dieudonné modules and supersingular abelian varieties revisited

This paper provides an expository account of Dieudonné modules and supersingular abelian varieties, offering simplified proofs for the uniqueness of products of supersingular elliptic curves and Oort's theorem on superspecial abelian varieties.

Original authors: Chia-Fu Yu

Published 2026-03-13

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Chia-Fu Yu

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are a detective trying to solve a mystery about a very special, hidden world of mathematical shapes called Abelian Varieties. These are complex, multi-dimensional shapes that behave like doughnuts but exist in higher dimensions.

This paper is a guidebook written by mathematician Chia-Fu Yu. Its goal is to explain how to classify these shapes, specifically a rare and mysterious type called Supersingular varieties. To do this, the author uses a powerful tool called Dieudonné Modules.

Here is the breakdown of the paper using simple analogies:

1. The Problem: Too Many Shapes, Not Enough Labels

In the world of math, there are many different "Supersingular" shapes. Sometimes, two shapes look completely different on the outside, but deep down, they are actually the same. Other times, they look similar but are actually different.

For a long time, mathematicians (like Deligne, Ogus, and Shioda) knew a rule: If you take two or more "Supersingular Elliptic Curves" (think of these as the basic building blocks, like Lego bricks) and glue them together, the resulting shape is unique. It doesn't matter which specific bricks you picked; if you have the same number of them, the final structure is identical.

However, proving why this happens was hard. This paper offers a new, simpler way to prove it.

2. The Tool: The "Fingerprint Scanner" (Dieudonné Modules)

To solve the mystery, the author introduces a tool called a Dieudonné Module.

The Analogy: Imagine every complex shape (Abelian Variety) has a secret, invisible "fingerprint" or a "DNA sequence." This sequence is too hard to read directly.
The Solution: The Dieudonné Module is a translation device. It takes the complex, messy DNA of the shape and translates it into a clean, organized list of numbers and rules (a module).
Why it helps: Instead of trying to compare two giant, complex shapes, mathematicians can just compare their "fingerprints." If the fingerprints match, the shapes are the same.

3. The Two Types of Shapes: Ordinary vs. Supersingular

The paper explains that these shapes come in two main flavors:

Ordinary: These are the "normal" shapes. Their fingerprints are a bit messy and irregular.
Supersingular: These are the "special" shapes. Their fingerprints are perfectly symmetrical and rigid.

The author shows that for Supersingular shapes, the fingerprint is so rigid that it forces the shape to be built from specific, identical Lego bricks (Supersingular Elliptic Curves).

4. The Big Discovery: The "Lego Brick" Theorem

The paper revisits a famous theorem. It says:

"If you build a tower using $g$ Supersingular Lego bricks, it doesn't matter which specific bricks you start with. As long as you have $g$ of them, the final tower is always the same."

The author provides a simple, elementary proof for this.

The Old Way: It was like trying to prove the tower is the same by climbing every single brick and measuring it with a microscope.
The New Way (This Paper): The author uses the "fingerprint scanner" (Dieudonné Modules). He shows that the fingerprints of all these towers are identical. Therefore, the towers must be identical. It's like realizing that every tower made of these specific bricks has the exact same barcode.

5. The "A-Number": Measuring the "Supersingularity"

The paper introduces a concept called the a-number.

The Analogy: Think of the a-number as a "Superspecial Score."
If a shape has the maximum possible score (equal to its dimension), it is Superspecial. This means it is a perfect, unblemished stack of identical Lego bricks.
The paper proves that if a shape has this perfect score, it must be a stack of identical bricks. There is no other way to build it.

6. The Twist: When Things Get Weird (The "Case I and Case II")

The author also looks at what happens when the shape is almost perfect but not quite (a Supersingular surface with a lower score).

Here, the "fingerprint" isn't enough to guarantee the shape is unique.
The paper divides these "almost perfect" shapes into two categories (Case I and Case II).
- Case II: Even though they aren't perfect, they are still all the same.
- Case I: Here, the shape depends on a specific parameter (like a dial you can turn). Depending on where you set the dial, you get a slightly different shape.
The author calculates exactly how many different shapes exist for each setting of the dial.

Summary: Why Does This Matter?

This paper is like a master key. It takes a very difficult, high-level problem in number theory and geometry and solves it using a clear, step-by-step logic.

It simplifies the proof that Supersingular shapes are built from identical bricks.
It shows how to use Dieudonné Modules (the fingerprint scanner) to count exactly how many different versions of these shapes exist.
It connects the abstract "fingerprint" to the physical shape, helping mathematicians understand the "landscape" (moduli space) where these shapes live.

In short, the author says: "Stop trying to measure the whole mountain. Just look at the unique pattern of the rocks at the base, and you'll know exactly what the mountain looks like."

1. Problem Statement

The paper addresses two primary objectives within the arithmetic geometry of abelian varieties in characteristic $p$ :

Expository Gap: To provide a clear, elementary introduction to the theory of Dieudonné modules, specifically focusing on their application to supersingular abelian varieties. The author aims to supplement existing literature (such as Harashita's work and Karemaker's lecture notes) with proofs that are distinct from original sources and more accessible.
Classification and Uniqueness: To revisit and provide simplified, elementary proofs for fundamental theorems regarding the uniqueness of products of supersingular elliptic curves and the structure of superspecial abelian varieties. Specifically, the paper seeks to clarify the relationship between the isomorphism classes of supersingular abelian varieties and their endomorphism rings (local and global).

2. Methodology

The author employs a combination of classical Dieudonné theory, deformation theory, and arithmetic group theory:

Dieudonné Module Theory: The paper establishes the anti-equivalence between $p$ -divisible groups and $W(k)$ -free Dieudonné modules (where $W(k)$ is the ring of Witt vectors). It utilizes the Manin-Dieudonné Theorem to classify these modules over an algebraically closed field by their slopes (Newton polygons).
Elementary Proofs: The author provides a self-contained proof of the Manin-Dieudonné Theorem (Section 3) using $F$ -spaces and cyclic algebras, avoiding more complex machinery found in original texts.
Deformation Theory: The paper utilizes the Cartier-Dieudonné theory and the Serre-Tate theorem to analyze the deformation of abelian varieties. Specifically, it examines the "a-number" locus ( $a=g$ ) in the moduli space of polarized abelian varieties to prove finiteness results.
Arithmetic of Endomorphism Rings: The core methodology for the uniqueness results involves analyzing the local endomorphism ring $O_p = \text{End}(X) \otimes \mathbb{Z}_p$ . The author uses the reduced norm map ($Nr$) and strong approximation theorems to relate the set of isomorphism classes of varieties with a fixed $p$ -divisible group to double coset spaces of the form $G(\mathbb{Q}) \backslash G(\mathbb{A}_f) / G(\widehat{\mathbb{Z}})$ .

3. Key Contributions and Results

A. Foundations and Classification

Dieudonné Modules: The paper defines Dieudonné modules, $p$ -divisible groups, and the associated functors (contravariant and covariant). It clarifies the distinction between ordinary and supersingular elliptic curves via the structure of their Dieudonné modules ( $M_1$ vs. $M_2$ ).
Manin-Dieudonné Theorem: A detailed proof is provided showing that any Dieudonné module over an algebraically closed field decomposes into a direct sum of simple modules $N^{\lambda}_{k}$ determined by rational slopes $\lambda$ .

B. Supersingular Abelian Varieties

Equivalence of Definitions: The paper proves that an abelian variety $X$ is supersingular (defined by having a single slope of $1/2$ ) if and only if it is isogenous to a product of supersingular elliptic curves over the algebraic closure.
Theorem 22 (Oort): If $X$ is a $g$ -dimensional abelian variety with $a(X)=g$ (superspecial), then $X$ is isomorphic to a product of $g$ supersingular elliptic curves ( $X \simeq E_1 \times \dots \times E_g$ ).
Theorem 23 (Deligne, Ogus, Shioda): Any product of $g$ supersingular elliptic curves is isomorphic to any other product of $g$ supersingular elliptic curves. The author provides a new proof using the uniqueness of the $p$ -divisible group structure.

C. Uniqueness and Isomorphism Classes (The Main Novelty)

The paper introduces Lemma 27, which establishes a bijection between three sets for a supersingular abelian variety $X$ :

The set $\Lambda_X$ of isomorphism classes of varieties $X'$ such that $X'[p^\infty] \simeq X[p^\infty]$ .
The set of isomorphism classes of locally free right ideals of the endomorphism ring $\text{End}(X)$ .
The quotient group $\mathbb{Z}_p^\times / \text{Nr}(\text{End}(X)_p^\times)$ , where $\text{Nr}$ is the reduced norm.

Key Implications:

Product Uniqueness: If $X = E \times X_1$ (where $E$ is a supersingular elliptic curve), then $\Lambda_X$ is a singleton. This implies that the product of supersingular elliptic curves is unique up to isomorphism, regardless of the specific curves chosen.
Dimension 2 Analysis: For supersingular abelian surfaces ( $g=2$ $g = 2$ ) with $a(X)=1$ $a (X) = 1$ (non-superspecial), the paper distinguishes two cases based on the parameter $t \in \mathbb{P}^1(k)$ $t \in P^{1} (k)$ defining the isogeny:
- Case I: $t \notin \mathbb{F}_{p^4}$ . Here, $\Lambda_X \simeq \mathbb{F}_p^\times / (\mathbb{F}_p^\times)^2$ .
- Case II: $t \in \mathbb{F}_{p^4} \setminus \mathbb{F}_{p^2}$ . Here, $\Lambda_X$ is a singleton.
- This leads to the result that any two non-superspecial supersingular surfaces in Case II are isomorphic.

D. Determination by $p$ -divisible Groups

The paper connects the arithmetic condition $\text{Nr}(O_p^\times) = \mathbb{Z}_p^\times$ to the geometric question of whether an abelian variety is uniquely determined by its $p$ -divisible group.

Corollary 33: An abelian variety $X$ ( $g>1$ ) is determined by its $p$ -divisible group if and only if it is supersingular and the reduced norm of its local units covers all of $\mathbb{Z}_p^\times$ .
Proposition 34: For $g=2$ , this condition holds for all $p=2$ , and for $p>2$ only when the variety is superspecial or in Case II.

4. Significance

Pedagogical Value: The paper serves as a valuable resource for understanding the intricate link between Dieudonné modules and the geometry of supersingular varieties, offering proofs that are more elementary than the classical literature.
Arithmetic-Geometric Bridge: It explicitly translates the geometric problem of classifying isomorphism classes of abelian varieties into an arithmetic problem involving the reduced norm of local endomorphism rings.
Oort's Conjecture: The paper situates its results within the context of Oort's conjecture (regarding the automorphism groups of generic abelian varieties). It references recent proofs for $g=3, 4$ and highlights the role of the $p$ -divisible group in determining the uniqueness of the variety.
Moduli Space Insights: By analyzing the locus where the reduced norm condition holds, the paper contributes to understanding the stratification of the moduli space of principally polarized abelian varieties, particularly regarding the "a-number" stratification.

In summary, Chia-Fu Yu's note revisits the classification of supersingular abelian varieties, providing a streamlined theoretical framework and new, elementary proofs for the uniqueness of products of supersingular elliptic curves, while establishing a precise arithmetic criterion for when a supersingular variety is uniquely determined by its $p$ -divisible group.