All loop BMN state energies from matrices

TL;DR

This paper analyzes the energies of all loop BMN states using a matrix quantum mechanics model derived from N=4 SYM, showing agreement with the BMN spectrum and discussing implications for the Bethe Ansatz conjecture.

Contribution

It introduces a matrix quantum mechanics approach to compute all loop BMN state energies and compares results with the BMN spectrum, revealing new challenges for the Bethe Ansatz.

Findings

BMN spectrum matches the matrix model predictions to all orders in coupling

Eigenvalue distributions correspond to dual geometry on S^5

Identifies issues with the all loop Bethe Ansatz conjecture

Abstract

We study a quantum corrected SO(6) invariant matrix quantum mechanics obtained from the s-wave modes of the scalars of N = 4 SYM on S^3. For commuting matrices, this model is believed to describe the 1/8 BPS states of the full SYM theory. In the large N limit the ground state corresponds to a distribution of eigenvalues on a S^5 which we identify with the sphere on the dual geometry AdS_5x S^5. We then consider non-BPS excitations by studying matrix perturbations where the off-diagonal modes are treated perturbatively. To a first approximation, these modes can be described by a free theory of "string bits" whose energies depend on the diagonal degrees of freedom. We then consider a state with two string bits and large angular momentum J on the sphere. In the large J limit we use a simple saddle point approximation to show that the energy of these states coincides precisely with the BMN spectrum to all orders in the 't Hooft coupling. We also find some new problems with the all loop Bethe Ansatz conjecture of the N=4 SYM planar spin chain model.