# Engineering Efficient Self-Play Chess: Search, Replay, and Throughput Under Limited Compute

arXiv: https://arxiv.org/abs/2609.37447 · HTML: https://mlsanity.com/paper/2609.37447/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 6 of its 11 references that are in the map.
- First posted on arXiv 27 Sep 2026.

## Builds on (most cited first, 6 of 6)

- [Prioritized Experience Replay](https://mlsanity.com/paper/1511.05952/) (2015)
- [Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm](https://mlsanity.com/paper/1712.01815/) (2017)
- [Accelerating Self-Play Learning in Go](https://mlsanity.com/paper/1902.10565/) (2019)
- [Monte-Carlo Graph Search for AlphaZero](https://mlsanity.com/paper/2012.11045/) (2020)
- [Learning to Stop: Dynamic Simulation Monte-Carlo Tree Search](https://mlsanity.com/paper/2012.07910/) (2020)
- [Targeted Search Control in AlphaZero for Effective Policy Improvement](https://mlsanity.com/paper/2302.12359/) (2023)

## Abstract (arXiv)

How strong can an AlphaZero-style chess system become under limited training compute when its entire learning loop is engineered for efficiency? We train from random initialization through searched self-play on a single eight-GPU node for 2.5 days. The resulting 6.32-million-parameter model reaches 3,251 benchmark Elo [3,206, 3,297] at 100,000 searches per move (estimated at under five seconds of thinking time) against a fixed-node Stockfish 13 ladder. The run ingests 3.25 million completed games, involves an estimated 100 billion search simulations, and makes 836.6 million training presentations. We investigate search allocation, replay and restart-state selection, policy representation, progressive model sizing, quantized inference, and throughput engineering. Alongside the retained design, we document plausible alternatives that failed to improve the complete learning loop or did not justify their cost. The reported strength is a result of the integrated system, not an isolated Elo gain attributable to any single choice.