AI models
GemNet-T (OC20)
Training compute
7.3×10¹⁷ FLOP
Parameters
1.9M
Published
Jul 1, 2021
GemNet-T (OC20) is an AI model developed by Technical University of Munich (Germany), first published in July 2021. It works in the materials science domain, on tasks such as molecular simulation, molecular property prediction, molecular representation learning and atomistic simulations.
Training it took an estimated 7.3×10¹⁷ FLOP of compute (estimation method: hardware). The model has 1,900,000 parameters. Training ran on 1 NVIDIA GeForce GTX 1080 Ti for about 60 hours.
Access: Open weights (restricted use). Its weights are openly available. Epoch AI rates the confidence of this record as likely.
Full record
- Organization
- Technical University of Munich
- Country of organization
- Germany
- Domain
- Materials science
- Task
- Molecular simulation, Molecular property prediction, Molecular representation learning, Atomistic simulations
- Training compute
- 7.3×10¹⁷ FLOP
- Compute estimation method
- Hardware
- Parameters
- 1,900,000
- Training hardware
- NVIDIA GeForce GTX 1080 Ti
- Chips used
- 1
- Training time
- 60 h
- Training power draw
- 278 W
- Model accessibility
- Open weights (restricted use)
- Open weights
- Yes
- Epoch confidence
- Likely
More from Technical University of Munich
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