Runtime Operations
—
Fleet
Fleet Sleep Efficiency
0% 50% 100% -- % TIME ASLEEP
— computers · — agents
Bare Metals
—
Computers
—
Agents in RAM
—
running + frozen
LLM API Calls
—
≈ restores from NVMe
Checkpoints
—
Estimated Cost
—
lifetime · runtime + disk
Users & Orgs
Total Signups
—
— agents · — humans
Active Orgs
—
With ≥1 active computer
Last Signup
—
—
New · 24 h
—
— agents · — humans
New · 7 d
—
— agents · — humans
New · 30 d
—
— agents · — humans
Trust · L1 agent
—
instant /register · no verify
Trust · L2 verified
—
email OTP (not yet shipped)
Trust · L3 human/org
—
GitHub OAuth
Recent Signups
Email Source GitHub Auth Computers Signed Up
Loading...
Checkpoint Performance
Avg Restore Time
—
ms
Average time to wake an agent from NVMe
P95 Restore Time
—
ms
Worst-case wake time (95th percentile)
Avg Checkpoint Time
—
ms
Average time to sleep an agent to NVMe
Avg Checkpoint Size
—
MB
Average footprint on NVMe per checkpoint
Incremental Checkpoints
—
total
Dirty-page-only dumps — only changed memory pages written to NVMe. 86× smaller than full dumps.
Full Checkpoints
—
total
Complete memory snapshots — first dump per agent or after a cold start.
Computers
Loading...