Abstract
Emerging computer hardware often trades reliability for energy efficiency; here we show that large-language models (LLMs) can be trained to tolerate this unreliability, and that rather than degrading, their error resilience actually increases as they grow. Modified neural scaling laws inferred from 40,000 GPU-hours of training runs on simulated faulty digital hardware quantify this trend and suggest that models learn to compute within "good" error-correcting codes, whose relative overhead remains finite no matter how large the model gets. This finding leads us to conjecture that appropriately trained LLMs may be formally fault-tolerant; if true, running AI inference on low energy, faulty hardware may be a path to substantial energy savings over the status quo.
Keywords
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
McCourt, T., Fiete, I. R., & Chuang, I. L. (2026). Fault-tolerant foundation models. https://omanscience.com/en/articles/fault-tolerant-foundation-models
MLA 9
McCourt, Trevor, et al. "Fault-tolerant foundation models." https://omanscience.com/en/articles/fault-tolerant-foundation-models.
Chicago (author–date)
McCourt, Trevor, Ila R. Fiete, and Isaac L. Chuang. 2026. "Fault-tolerant foundation models." https://omanscience.com/en/articles/fault-tolerant-foundation-models.
Harvard
McCourt, T., Fiete, I. R. and Chuang, I. L. (2026) 'Fault-tolerant foundation models', Available at: https://omanscience.com/en/articles/fault-tolerant-foundation-models.
Vancouver
McCourt T, Fiete IR, Chuang IL. Fault-tolerant foundation models. https://omanscience.com/en/articles/fault-tolerant-foundation-models
IEEE
T. McCourt, I. R. Fiete, and I. L. Chuang, "Fault-tolerant foundation models," https://omanscience.com/en/articles/fault-tolerant-foundation-models.