A second mining rig: what actually changes when you scale
The first rig runs, revenue drips in, and the hands itch for a second one. Fair enough — but a second rig is not "the first one times two". Some things scale for free, some hit the hard physical limits of your home, and some turn into a separate job. Let's go through it in order, so rig number two brings revenue instead of doubled problems.
The short version (for the no-time-to-read)
- Wiring scales the worst. A home has a physical kilowatt ceiling — learn yours before buying cards, not after.
- Stabilize the first rig first. A copy of an unstable rig is two unstable rigs and twice the midnight failures.
- Knowledge and presets scale for free: a second rig on the same cards gets configured several times faster than the first.
- Heat and noise double completely. If one rig is already "barely tolerated", two will fit neither the room nor the family.
- Spare parts start paying off: with two rigs, keeping a spare riser, cable and fan stops being paranoia and becomes normal.
What scales badly: wiring, heat, noise
Three limits that money spent on cards cannot move:
- The grid limit. A standard home connection holds a continuous load of a few kilowatts. One 1.5 kW rig is fine; two already push household wiring to the edge, especially with a water heater and AC. Work out the budget with our wiring article and run a dedicated circuit before buying, if needed.
- Heat. Two rigs are two 1.5 kW heaters. In summer a room without exhaust becomes uninhabitable (solutions in the summer article).
- Noise. Decibels do not add linearly, but two rigs are almost twice as audible. If rig one lives in a living room, rig two will not fit there — plan the location early (options in the relocation article, noise in the quiet article).
What scales well: knowledge, presets, spares
The pleasant side of scaling — almost everything intellectual is free:
- presets. An overclock is tuned to the card model and memory, not bound to one specific card forever: a second rig on the same cards starts from proven working points. Our catalog is exactly about that;
- diagnostic experience. You will dissect a midnight failure on the second rig in minutes — the checklists exist already (from hangs to risers);
- spare parts. One spare riser and a couple of cables now insure two rigs — the insurance cost per rig halves;
- infrastructure. The router, monitoring and tariff calculations are shared across the whole operation.
The pre-second-rig checklist
- Rig one stable for a month+ without manual intervention. If not — finish it first: doubling chaos is easy, untangling it costs twice as much.
- Wiring calculated: total load with headroom, circuits and breakers known, outlets checked.
- Location solved: where the second rig's heat and noise go — decided before the purchase.
- Management untied from your presence: reboots, preset changes, notifications — all must work remotely, or two rigs will chain you to the house.
- Revenue counted honestly: at your tariff and current difficulty (on network difficulty and how to count).
Where automation fits
The manual approach scales worst of all: two rigs by hand means twice the tuning, twice the checks, twice the midnight drives. OCBot installs per rig, and everything is managed from one Telegram: overclock tuning, failure handling, notifications — per farm, separately. Good presets travel between rigs, so the second one gets configured several times faster than the first — that is what proper scaling looks like.
This article is general information summarizing public practice; actual decisions depend on your wiring, space and tariff. OCBot is an independent project: not affiliated with NVIDIA, HiveOS or Hiveon. Use at your own risk — see the disclaimer.