01 / The mechanism and its boundary
What is being described
FLOP accounting estimates or verifies the total number of floating-point operations (FLOP) that a training run or other workload performs, usually to compare it with a threshold set by a rule S-0029 S-0053.
Shavit lists total training compute among the rules a verifier might enforce, noting that it has proven to be an indicator of model capabilities S-0029. US Executive Order 14110 required reporting for models trained with more than 10^26 operations S-0053 until its revocation in January 2025 S-0069, and a proposed international agreement sets a prohibited threshold of 10^24 FLOP and a monitored threshold of 10^22 FLOP S-0063. The EU AI Act presumes that a general-purpose AI model trained with more than 10^25 FLOP has high-impact capabilities, with notification of the European Commission required since 2 August 2025 S-3542 S-3543, and California's SB 53 defines frontier models by more than 10^26 operations of training compute S-3544. Proposed ways to count or cap FLOP include:
- Hardware time. Shavit converts a FLOP threshold into accelerator-days by assuming that every accelerator runs at its full rate with perfect parallelization, a conservative assumption that gives the fewest accelerator-days a run of that size could occupy S-0029.
- Energy. Wasil and colleagues suggest that a data centre's measured energy use could be converted into an approximate FLOP count S-0062.
- Telemetry. RAND lists estimating a workload's model FLOPs utilization (MFU) and physical signature, such as power, as a research problem S-0002, one that bears on on-chip telemetry.
- On-chip budgets. Offline licensing ties chip use to a renewable licence carrying a compute budget S-0057, as in hardware performance throttling and licensing.
Sastry and colleagues call compute a good high-level proxy for the risk of general-purpose frontier systems, though not necessarily of some narrow ones, and expect thresholds to need changing as algorithms and hardware improve S-0053. One system overview argues that coarse totals such as FLOP counts are not enough and seeks evidence about individual workloads S-0018.
Connections in the research map
Related research
Sources and provenance
- S-0029 / Tier B
What does it take to catch a Chinchilla? Verifying Rules on Large-Scale Neural Network Training via Compute Monitoring ↗
Y. Shavit · 2023 · arXiv
Supports: total training compute as a rule and an indicator of capabilities; a threshold of H FLOPs converted to chip-days using each chip's FLOPs per day at full, perfectly parallel use
Locator: §2.1; §3.2, Table 1
Version and catalogue details - S-0053 / Tier B
Computing Power and the Governance of Artificial Intelligence ↗
G. Sastry, L. Heim, H. Belfield, M. Anderljung, M. Brundage, J. Hazell, C. O'Keefe, G. K. Hadfield, R. Ngo, K. Pilz, G. Gor, E. Bluemke, S. Shoker, J. Egan, R. F. Trager, S. Avin, A. Weller, Y. Bengio, D. Coyle · 2024 · arXiv
Supports: EO 14110 reporting threshold of 10^26 operations; compute as a good high-level proxy for risk of general-purpose frontier systems; thresholds need changing with algorithmic and hardware progress
Locator: thresholds; limitations
Version and catalogue details - S-0063 / Tier B
An International Agreement to Prevent the Premature Creation of Artificial Superintelligence ↗
A. Scher, D. Abecassis, P. Barnett, B. Abeyta · 2025 · Machine Intelligence Research Institute
Supports: Strict Threshold 10^24 FLOP and Monitored Threshold 10^22 FLOP
Locator: §4
Version and catalogue details - S-3542 / Tier A
Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act) ↗
European Parliament, Council of the European Union · 2024 · Official Journal of the European Union, OJ L, 2024/1689
Supports: EU AI Act 10^25 FLOP presumption, notification, application from 2 August 2025
Locator: Arts 51(2), 52(1), 113(b)
Version and catalogue details - S-3543 / Tier A
Guidelines on the scope of the obligations for general-purpose AI models established by Regulation (EU) 2024/1689 (AI Act) ↗
European Commission · 2025 · European Commission, Communication C(2025) 5045 final
Supports: 10^25 FLOP presumption and notification; entry into application on 2 August 2025
Locator: §2.3.1–2.3.2; landing page
Version and catalogue details - S-3544 / Tier A
California Senate Bill 53 (2025): Transparency in Frontier Artificial Intelligence Act ↗
California State Legislature · 2025 · Statutes of 2025, Chapter 138 (Business and Professions Code §22757.10 et seq.)
Supports: frontier model defined by more than 10^26 integer or floating-point operations
Locator: §22757.11(i)
Version and catalogue details - S-0062 / Tier B
Verification methods for international AI agreements ↗
A. R. Wasil, T. Reed, J. W. Miller, P. Barnett · 2024 · arXiv
Supports: facility energy estimates could be converted into an approximation of FLOPs
Locator: Energy monitoring
Version and catalogue details - S-0002 / Tier B
Verifying International Agreements on AI: Six Layers of Verification for Rules on Large-Scale AI Development and Deployment ↗
M. Baker, G. Kulp, O. Marks, M. Brundage, L. Heim · 2025 · RAND Corporation
Supports: estimating MFU and physical signature (e.g. power) as a research problem
Locator: Table 2, Appendix A.6
Version and catalogue details - S-0057 / Tier B
Hardware-Enabled Governance Mechanisms: Developing Technical Solutions to Exempt Items Otherwise Classified Under Export Control Classification Numbers 3A090 and 4A090 ↗
G. Kulp, D. Gonzales, E. Smith, L. Heim, P. Puri, M. J. D. Vermeer, Z. Winkelman · 2024 · RAND Corporation
Supports: offline licensing with a renewable licence carrying a compute budget
Locator: p. viii
Version and catalogue details - S-0018 / Tier B
A System Overview for Near-Term, Low-Trust AI Compute Verification ↗
N. Cankaya · 2026 · Machine Intelligence Research Institute
Supports: coarse metrics like total FLOPs insufficient; per-workload evidence sought
Locator: verification goals
Version and catalogue details - S-0069 / Tier A
Executive Order 14148: Initial Rescissions of Harmful Executive Orders and Actions ↗
Executive Office of the President · 2025 · Federal Register, 90 FR 8237 (document 2025-01901, published 2025-01-28)
Supports: revocation of EO 14110 on 20 January 2025
Locator: Sec. 2(ggg)
Version and catalogue details
- Source review date
- 2026-09-25
- Drafted by (source map)
- ai
- Review handles (source map)
- codex-review