In the world of mining, the race for hashrate often overshadows the main goal — pure profit. Your primary operational expense, which eats up the lion’s share of revenue, is your electricity bill. Undervolting is not just a technical trick, but a fundamental business strategy that allows you to attack this cost head-on. It’s the art of making chips work not louder, but smarter, finding their peak efficiency.
We will not sacrifice performance. We will optimize it. The goal is not just to reduce consumption, but to find that “sweet spot” where every watt spent brings maximum hashes, and thus, maximum money into your pocket.
Why Undervolting is a Winning Strategy
Understand this: manufacturers sell you an averaged product. It must work both in a cold basement in Siberia and in a hot garage in Texas. Therefore, the factory voltage settings are always set high “with a margin.” This safety margin is your potential for optimization.
By reducing voltage, we achieve several goals at once:
- Radical reduction in power consumption. This is the direct and most obvious benefit. A 20-30% reduction in consumption with only a 5-10% loss in hashrate is an absolutely realistic scenario that dramatically changes the economics of mining.
- Improved J/TH (Joules per Terahash). This is a key metric. If your ASIC consumes 3000 W at 100 TH/s, its efficiency is 30 J/TH. If, after undervolting, it delivers 92 TH/s at 2200 W, its efficiency becomes ~23.9 J/TH. You have become 20% more efficient. You are mining cryptocurrency cheaper.
- Reduced temperature and noise. Less voltage means less heat. Chips run cooler, and fans spin slower. Your ASIC will last longer and be less irritating with noise.
- Reduced load on wiring and PSU. Operating in a gentle mode reduces risks associated with electrical grid overload and power supply failure.
From Theory to Practice: A Step-by-Step Undervolting Process
Enough theory. Let’s get to business. The process requires methodicalness and patience, but it’s not complicated. We will proceed step by step.
Step 0: Preparation — The Foundation of Success
Before changing anything, you need to establish a baseline and prepare your tools. Ignoring this step is a direct path to incorrect conclusions.
What you will need:
- Custom firmware. Stock software will not give you the necessary control. Install proven firmware like Braiins OS+ or Vnish. They provide access to fine-tuning voltage and frequencies.
- Physical wattmeter. Forget the figures in the web interface — they often lie. Only a wattmeter installed between the outlet and the ASIC’s power supply will show you the real power consumption of the entire system. This is your most important instrument.
- Stable environment. Ensure that the room temperature is constant. Any external changes will affect the purity of the experiment.
- Logbook for records. A simple Excel spreadsheet or notebook. You will meticulously record the results of each step here.
Measure baseline parameters:Turn on the ASIC with factory settings. Let it run for 30-40 minutes to stabilize. Record the following in your log:
- Real power consumption from the wattmeter (W).
- Average hashrate (TH/s) — check it not in the miner interface, but on your pool’s statistics page. This is the most honest indicator.
- Chip temperature for each board (°C).
- Number of hardware errors (HW Errors). Ideally, there should be 0.
Now you have a baseline against which you will compare all subsequent results.
Step 1: First Power Reduction
Open the web interface of your custom firmware. Find the tuning section (Tuner). Instead of directly adjusting voltage, most modern firmware allows you to control the total power consumption (Power Limit). This is safer and more intuitive.
Reduce the set power by 100-200 W from your baseline. For example, if the wattmeter showed 3100 W, set the limit to 2900 W. Do not touch the frequency. Apply the settings.
Step 2: Analysis and Recording of Results
Now for the most important part — observation. Let the miner run in the new mode for at least 20-30 minutes. It must be fully stabilized. After that, record all the same parameters in your log again:
- Power consumption from the wattmeter.
- Average hashrate from the pool.
- Chip temperature.
- Number of HW errors.
Now calculate the new J/TH value (Watts / Terahashes) and compare it with the baseline. It should have decreased. You are on the right track.
Step 3: Iterations — Finding the “Sweet Spot”
Repeat Step 1 and Step 2. Reduce the power limit in small increments — by 50-100 W. Each time, wait for stabilization and carefully record the results. You will notice that consumption drops much faster than hashrate. Your energy efficiency (J/TH) will improve with each step.
This methodical process is the search for that “sweet spot” — the point where J/TH is minimal.
Step 4: How to Know When to Stop
At some point, you will reach a limit. Signs that you have gone too far:
- Sharp drop in hashrate. Hashrate begins to drop disproportionately compared to the power reduction.
- Appearance of HW errors. This is the main red flag. The chips are not receiving enough power for stable calculations.
- Instability. The miner starts spontaneously rebooting, or one of the boards “drops out.”
As soon as you encounter one of these signs, you have hit rock bottom. Revert to the last stable configuration, slightly increasing the power (by 25-50 W) for a safety margin. Congratulations, you have found your ideal undervolting profile.
Live Example: Antminer S19j Pro Undervolting
Let’s look at the numbers. Let’s take the popular S19j Pro and an electricity rate of $0.08 per kWh.
Factory Settings (Baseline):
- Consumption: 3100 W
- Hashrate on pool: 100 TH/s
- Efficiency: 31 J/TH
- Daily electricity costs: 3.1 kW * 24 h * $0.08 = $5.95
After proper undervolting (found “sweet spot”):
- Consumption: 2350 W (24% reduction)
- Hashrate on pool: 92 TH/s (only 8% reduction)
- Efficiency: 25.5 J/TH (17.7% improvement)
- Daily electricity costs: 2.35 kW * 24 h * $0.08 = $4.51
Result: You sacrificed 8% of gross hashrate but save $1.44 per day or about $43 per month per device. You didn’t just reduce costs — you increased net profit and made your business more resilient to cryptocurrency price drops.
Conclusion: Undervolting is not a Compromise, but Optimization
Undervolting is a mandatory procedure for any miner who counts money. It’s a shift from a “more terahashes” mindset to a “more net profit” mindset. You get cooler, quieter, more durable, and most importantly, more profitable equipment.
By spending a few hours on methodical tuning, you secure a competitive advantage for months and years to come. You transform your standard ASIC into a miner perfectly calibrated for your personal economy.
