The mining winter: what to do when profit turns negative

It happens to everyone: the price rolls down, per-card revenue falls below its electricity bill, and the farm turns from a money printer into a space heater you pay for. No secrets needed here — winter handled by three things: honest arithmetic, lower consumption, and knowing when the rig should sleep. In order, without panic.

The short version (for the no-time-to-read)

  • Count first, panic later. The number on the pool screen is revenue, not profit. Compute against your own tariff (formulas and a calculator in our article).
  • A negative at max overclock does not mean shutting down just yet. Cut watts first: a card dropped from 280 to 180 W earns nearly the same coins and pays far less for itself.
  • The same card can be profitable in efficiency mode and unprofitable in hashrate mode. The OC style changes economics more than it seems.
  • If even the lean preset loses money, the rig should sleep. Stopping is a legitimate mode, not a project failure (more on stopping here).
  • Selling hardware at the bottom is the worst option. Prices sag and recover in cycles; panic lasts weeks. Do the upgrades you cannot do under load — and wait for the window.

Step 1. Honest arithmetic: revenue is not profit

The classic winter mistake is staring at the revenue figure on the pool screen. The correct line is:

daily profit = daily revenue − daily electricity, where electricity is the sum of all cards' watts × 24 hours × your tariff. The tariff matters more than any forecast: at $0.07/kWh, tens of watts decide everything.

Example: a card earns 150 units/day in coins and draws 250 W. At $0.07/kWh electricity costs about 0.42 units/day, profit 149.6. Price halves: revenue 75, profit ~74.5 — still fine. Tariffs at 0.3+ $/kWh change the story completely: the same rig pays 1.8/day and at halved revenue is already negative. Note: not the whole farm dies first — its hungriest half does.

Step 2. Cut the watts before switching the rig off

By winter a rig usually runs in hashrate mode: max clocks, max draw, max hashrate — summer logic. But winter revenue is measured in coins per watt-hour, and the economics flip:

In practice: pick a "best coins per watt" preset per card and recount profit on it. Often that alone turns winter into "quieter, fewer watts" instead of a loss. Holding back the watts of neighboring cards to fit the PSU budget is also legitimate: the shared power budget belongs to the profitable cards.

Step 3. When the rig should sleep — and why it is not the end

If even after the efficiency preset the card eats more than it earns, stopping is justified. Practical notes from public community experience:

Where automation fits

While cards are in the black, the bot's work is ordinary. Winter is different in one respect: a hashrate-tuned rig asks for a fresh OC sweep at every price move. OCBot keeps each card in the chosen style (compromise / efficiency / hashrate), respects the rig's shared power budget, and quickly returns to the last good preset after algorithm changes. Whether to run or sleep stays your call — the bot shows the numbers, it does not make strategic decisions for you.