Pretraining at scale
Large models trained on broad data, the base every other capability builds on.
capabilityBuilt today Open on the map →
Facts
- Lane
- Algorithms and learning
- Wave
- 3 of 12 (build order, not a date)
- Needs
- 6 boxes all the way back through hard links
- Unlocks
- 100 boxes downstream through hard links
- Leverage
- None: already built
- Cluster
- Science and medicine
Evidence as of October 2026
Epoch AI finds training compute can keep growing through 2030, with power, chips, data and latency as the four limits.