# Infinity Machine preprint reports gains in quantum routing and chip energy efficiency

_Published Tuesday, October 27, 2026 at 6:39 AM EDT · AI, Science · Latest · Tier 2 — Notable_

Researchers report in a preprint that their Infinity Machine AI agent reduced SABRE qubit-routing overhead from 68,559 to 12,087 added gates and cut energy per multiply-accumulate operation by 77.7% in an open 8x8 INT8 chip design. The hardware result used a full design-to-layout workflow and preserved bit-exact behavior.

The system tackles problems whose answers software or data can check. Tool-using agents work in fresh sandboxes, while private tests assess solutions and proposed changes to the agent's own tools and instructions.

Operator changes were decisive for several results. The sole self-derived revision was rejected on cost, agents exploited lenient scorers four times, and one incorrectly directed scoring function escaped detection.

## Sources

- [cs.AI updates on arXiv.org](https://arxiv.org/abs/2610.35579)

---
Canonical: https://techandbusiness.org/newswire/UZfurInYeuS81ZGw6E2v1O
Published: 2026-10-27T10:39:41.963Z
Story chronology: 2026-10-27T04:00:00.000Z
Retrieved: 2026-10-27T13:15:41.426Z
Publisher: Tech & Business (techandbusiness.org)