1: \begin{abstract}

2: \noindent

3: While semidefinite programming (SDP) has traditionally been limited to moderate-sized problems, recent algorithms augmented with matrix sketching techniques have enabled solving larger SDPs.

4: However, these methods achieve scalability at the cost of an increase in the number of necessary iterations, resulting in slower convergence as the problem size grows.

5: Furthermore, they require iteration-dependent parameter schedules that prohibit effective utilization of warm-start initializations important in practical applications with incrementally-arriving data or mixed-integer programming.

6: We present SpecBM, a provably correct, fast and scalable algorithm for solving massive SDPs that can leverage a warm-start initialization to further accelerate convergence.

7: Our proposed algorithm is a spectral bundle method for solving general SDPs containing both equality and inequality constraints.  

8: Moveover, when augmented with an optional matrix sketching technique, our algorithm achieves the dramatically improved scalability of previous work while sustaining convergence speed.

9: We empirically demonstrate the effectiveness of our method, both with and without warm-starting, across multiple applications with large instances.

10: For example, on a problem with 600 million decision variables, SpecBM achieved a solution of standard accuracy in less than 7 minutes, where the previous state-of-the-art scalable SDP solver requires more than 16 hours.  Our method solves an SDP with more than $10^{13}$ decision variables on a single machine with 16 cores and no more than 128GB RAM; the previous state-of-the-art method had not achieved an accurate solution after 72 hours on the same instance.

11: We make our implementation in pure JAX publicly available\footnote{\url{https://github.com/rangell/specbm}}.

12: \end{abstract}

13: