AI models
MEGNet (crystal band gap model)
Training compute
4.5×10¹⁷ FLOP
Parameters
26.1K
Published
Apr 10, 2019
MEGNet (crystal band gap model) is an AI model developed by University of California San Diego (United States), first published in April 2019. It works in the materials science domain, on tasks such as molecular property prediction.
Training it took an estimated 4.5×10¹⁷ FLOP of compute. The model has 26,128 parameters. It was trained on roughly 36.7K datapoints. Training ran on 1 NVIDIA GeForce GTX 1080 Ti for about 28 hours.
The reference paper has 1,301 citations. Epoch AI rates the confidence of this record as likely.
Full record
- Organization
- University of California San Diego
- Country of organization
- United States
- Domain
- Materials science
- Task
- Molecular property prediction
- Training compute
- 4.5×10¹⁷ FLOP
- Parameters
- 26,128
- Dataset size
- 36.7K
- Training hardware
- NVIDIA GeForce GTX 1080 Ti
- Chips used
- 1
- Training time
- 28 h
- Training power draw
- 283 W
- Citations
- 1,301
- Epoch confidence
- Likely
More from University of California San Diego
SourceEpoch AI, 'AI Models'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/ai-models. Licensed under CC BY 4.0.
← All ai models