[Paper Review] Noncommutative Geometry and Matrix Theory: Compactification on Tori

Alain Connes, Michael R. Douglas+1 more|arXiv (Cornell University)|Nov 20, 1997

Noncommutative and Quantum Gravity TheoriesPhysics and Astronomy6 references1,145 citations

TL;DR

This paper proposes that compactification of Matrix theory on noncommutative tori—formulated using noncommutative geometry—yields new solutions to M-theory compactifications with constant background three-form fields. By generalizing toroidal compactifications via Moyal-deformed gauge theories, it identifies an $SL(2,\mathbb{Z}) \times SL(2,\mathbb{Z})$ duality symmetry in the moduli space of constant curvature connections, linking noncommutative geometry to M-theory and T-duality in type II string theory.

ABSTRACT

We study toroidal compactification of Matrix theory, using ideas and results of non-commutative geometry. We generalize this to compactification on the noncommutative torus, explain the classification of these backgrounds, and argue that they correspond in supergravity to tori with constant background three-form tensor field. The paper includes an introduction for mathematicians to the IKKT formulation of Matrix theory and its relation to the BFSS Matrix theory.

Motivation & Objective

• To generalize toroidal compactifications of Matrix theory using noncommutative geometry, extending standard commutative torus backgrounds.
• To establish a correspondence between noncommutative torus compactifications and supergravity solutions with constant background three-form fields.
• To demonstrate that the moduli space of constant curvature connections on noncommutative tori describes space-time in the BFSS model, analogous to conventional compactification.
• To identify a new $SL(2,\mathbb{Z}) \times SL(2,\mathbb{Z})$ duality symmetry in the compactified theory, linking to T-duality in type II string theory.
• To provide a physical interpretation of noncommutative compactifications as arising from M-theory compactified on a torus with a light-like circle and non-zero three-form flux.

Proposed method

• Formulates the IKKT matrix model in Euclidean signature with a holomorphic action functional involving 10 Hermitian matrices $X_i$ and 16 Weyl spinors $\Psi^\alpha$, using an invariant inner product and Dirac matrices.
• Introduces a deformation of the gauge theory Lagrangian using Moyal brackets instead of Poisson brackets, encoding noncommutativity via a two-form parameter $\theta_{ij}$.
• Identifies the defining relations of toroidal compactification as equivalent to the definition of a connection on the noncommutative torus, linking noncommutative geometry to matrix theory.
• Constructs two commutative tori from a noncommutative torus—one for odd cohomology (space-time) and one for even cohomology (dual moduli space), each carrying an $SL(2,\mathbb{Z})$ action.
• Derives a physical interpretation where the noncommutative torus compactification corresponds to M-theory on a torus with a constant background three-form potential $C_{ij-}$, particularly in the case of a light-like circle.
• Uses the BPS mass formula and string world-sheet description to verify that the proposed $SL(2,\mathbb{Z}) \times SL(2,\mathbb{Z})$ symmetry is consistent with T-duality and supersymmetry.

Experimental results

Research questions

• RQ1 How can toroidal compactifications of Matrix theory be generalized beyond the commutative torus using noncommutative geometry?
• RQ2 What is the physical interpretation of noncommutative torus compactifications in terms of M-theory and supergravity backgrounds with constant three-form flux?
• RQ3 Does the noncommutative torus compactification exhibit an $SL(2,\mathbb{Z}) \times SL(2,\mathbb{Z})$ duality symmetry, and if so, how does it relate to T-duality in type II string theory?
• RQ4 Can the moduli space of constant curvature connections on the noncommutative torus be identified as space-time in the BFSS matrix model?
• RQ5 What is the role of the two-form parameter $\theta_{ij}$ in deforming the gauge theory and how does it correspond to a background three-form field in M-theory?

Key findings

• The defining relations of toroidal compactification in the BFSS/IKKT model are mathematically equivalent to the definition of a connection on the noncommutative torus, establishing a precise link between matrix models and noncommutative geometry.
• The moduli space of constant curvature connections on the noncommutative torus—associated with odd cohomology—serves as space-time in the compactified matrix theory, generalizing the standard torus compactification.
• Two distinct $SL(2,\mathbb{Z})$ duality groups act on the Teichmüller space of the noncommutative torus: one on the odd cohomology (space-time) and one on the even cohomology (dual moduli space), leading to a $SL(2,\mathbb{Z}) \times SL(2,\mathbb{Z})$ symmetry.
• The noncommutative compactification corresponds to M-theory compactified on a torus with a constant background three-form potential $C_{ij-}$, particularly when one dimension is null, and this background is consistent with supersymmetry.
• The BPS mass formula and string world-sheet description respect the $SL(2,\mathbb{Z}) \times SL(2,\mathbb{Z})$ duality symmetry, providing physical evidence for its validity in the matrix theory framework.
• The paper conjectures that gauge theory on the noncommutative torus is the correct description of the deformed matrix theory, offering a concrete, perturbatively defined framework for studying large $N$ limits and duality symmetries.