Stabalize Calibration network (calibnet)

[BigLep]

Background

At least as of 2025-11-25, one SP has more than 50% of the power.

If that SP is temporarily down, the time to get messages on the chain takes longer. (This happened on 2025-10-28.)

We want to have enough power here that it’s non-trivial for non-implementation teams to fork the calibration network.

Done Criteria

3 "just right" sized miner from 3 different organizations so there is enough of a moat while having some resiliency to one SP being down.

Specifically, this translates to:

• Venus keep their miner t04040 at ~250 TiB of QAP
• Curio reducing miner t0143103 to 250 TiB of QAP
• Chainsafe bringing up a miner with 250 TiB of QAP. (Curio should be able to help seal 25 TiBs of sectors.)

As a bonus, since this is about calibnet stability, we'd like to have Probelab add monitoring for calibnet to https://probelab.io/filecoin

Why Important

This is coming at a time where there is a lot of development and testing on calibration with Filecoin Onchain Cloud.

Notes

1. This was discussed at FDS-7 (session notes)
2. Slack thread in #fil-net-calib-discuss

[BigLep]

Ccing @ianconsolata and @lanzafame because it was mentioned in FDS-7 that you were interested in being kept in the loop on this topic.

[ianconsolata]

Thanks! Will you coordinate directly with the Venus and Chainsafe teams on this, or do you want us to bring it up in our meetings with them?