Pod downtime, trailing 7 days

Osprey and West TX TBD (non-producing buildout sites) excluded. Source: InfluxDB MARAGON_Hashrate, 2026-08-15 16:20 UTC – 2026-08-22 16:20 UTC.

Method: pod hashrate averaged into 5-minute buckets; a bucket counts as "down" when mean hashrate < 0.1 TH/s (i.e. the pod produced essentially nothing — gen down, EMS/sleep, or connectivity loss all show up this way; this does not distinguish cause, and does not count partial degradation where some miners in a pod stay up while others are offline). ~2,000 of ~2,016 possible 5-min buckets per pod were present (small gaps from collector restarts, not downtime). Cross-checked against gen_down_events for the same sites/groups — same pods (Nate 1, GN, WW, Will 3) rank highest there too. Dan 1 is showing 0 hashrate as of this snapshot (2026-08-22 ~15:45 UTC) — currently down, not just a historical figure.