IOTA says its 16.2B Project Orion run is training across 187 scattered GPUs
bittingthembits · x · 2026-08-04
Macrocosmos/IOTA says its Project Orion run is training a 16.2B-parameter model across scattered infrastructure rather than a single cluster.
The screenshot shows the run handling up to 256 machines, with about 187 active nodes online. It has already processed 50.9B tokens at roughly 62.6K tokens/s, and the loss curve continues to fall. The pitch is that independently operated GPUs in different places can be orchestrated like one reliable supercomputer, so idle compute anywhere can be used for training.
More from Infra
- K3 looks stronger than expected, and the argument is to own your inference stack — hsu_byron · 2026-08-04
- Photon 2.0 compiles Moondream, Qwen 3.5 and Gemma 4 into megakernels — sloppenheimer · 2026-08-04
- Minimax H3 runs on an RTX 3060 12GB, but a 10-second clip takes 30 minutes — irmemon225 · 2026-08-04
- Minimax H3 generation times on a 5070 Ti compare ComfyUI, Sage Attention, and Easy Cache — TheRedHairedHero · 2026-08-04
- Open-source H3Zero wraps MiniMax H3 in a Modal-hosted UI with API support — LegacyV1 · 2026-08-04
- MiniMax H3 already handles 15-second clips, but one RTX 5090 run took over 6 minutes — Careless-Constant-33 · 2026-08-04