---
title: Ground-State Statistics of Sparse Max-Cut at 10⁸ to 10¹⁰ Spins: Extrapolations Bracket the Parisi Constant | SpinForge
url: https://spin-forge.com/proof/p0
description: 0.738–0.776: extrapolated (c − ½)√d at infinite degree from converged runs; Parisi P* = 0.7632. Converged anneals at 10⁸ to 10¹⁰ spins on 8×H100, with…
---

RESEARCH RECORD · P0

# Ground-State Statistics of Sparse Max-Cut at 10⁸ to 10¹⁰ Spins: Extrapolations Bracket the Parisi Constant

Converged anneals at 10⁸ to 10¹⁰ spins on 8×H100, with independent CPU recounts at 10¹⁰ and overlaps between independent solutions computed over all 10⁹ spins.

0.738–0.776 extrapolated (c − ½)√d at infinite degree from converged runs; Parisi P* = 0.7632

0.00011 pp standard deviation of the cut over 15 runs at N = 10⁹ (79.5954–79.5958%)

h = 0.49997 between two equal-cut solutions of one 10¹⁰-spin instance

2.66 pp annealed vs greedy-quench branch at N = 10⁹, both recounted

- (c − ½)√d rises from 0.683 (d = 3) to 0.745 (d = 16); extrapolations bracket Parisi: 0.738–0.776 vs 0.7632
- Self-averaging: sd 0.00011 pp (15 runs, 10⁹); 0.0027 pp (4 instances, 10¹⁰)
- Equal cuts from uncorrelated configurations (h = 0.49997 at 10¹⁰)
- Annealed 79.596% vs greedy-quench 76.937% at 10⁹; the cooling rate selects the branch
- Planted structure recovered above a sharp, size-independent f_c = 0.488–0.496 (N = 2k to 128k)
[Request the full paper by email](https://spin-forge.com/contact?product=portfolio&brief=Please%20send%20the%20full%20paper%3A%20Ground-State%20Statistics%20of%20Sparse%20Max-Cut%20at%2010%E2%81%B8%20to%2010%C2%B9%E2%81%B0%20Spins%3A%20Extrapolations%20Bracket%20the%20Parisi%20Constant%20%28P0%29.%0A%0AOur%20organization%20and%20intended%20use%3A%0A)

## More research records

- [Max-Cut at 16.5 Billion to 200 Billion Variables on One 8×H100 Node](https://spin-forge.com/proof/p2)
- [Time-to-Target Scaling on D-Wave's Pegasus Instances: Exponent 0.44 vs 1.69 for Quantum Annealing with Error Correction](https://spin-forge.com/proof/p1)

[All results, charts and papers](https://spin-forge.com/proof)

Results describe the stated inputs, hardware and conditions and are not a guarantee of results in your environment. See the [Terms of Use](https://spin-forge.com/terms).