IOTA Overview

IOTA tackles one of the most challenging and important issues of our time - the training of massive AI models. Our philosophy is to build in public and iterate quickly. This means both transparent experiments, and permissionless participation.

IOTA coordinates global compute to create an indistinguishable compute fabric from distributed, hetereogenous nodes. This unlocks truly collaborative open models, and a new paradigm in cost-efficient training.

Snapshot
Completed Experiments
0
Over 6 months
Train at Home App Downloads
0
Since launch
Total Tokens Processed
0.0T
Across all experiments

Highlighted experiments

Training Loss

Select individual experiments or view all at once. Loss curves for highlighted experiments. Select runs below to compare.

Early Training Loss

Median loss at 50M tokens. Lower is better.

Throughput

Tokens per replica second (TPRS). Higher is better.

Hardware
GPU Distribution

Train at Home hardware breakdown across contributing compute nodes.

Geographical contribution
Compute Node Locations
Top locations
Paris
0
Los Angeles
0
New York
0
San Francisco
0
Chicago
0
Mumbai
0
Rio de Janeiro
0
London
0
Berlin
0
Tokyo
0
Moscow
0
Singapore
0
Dubai
0
Sydney
0
Cape Town
0
0
Nodes
0
Locations