Bitcoin mining is the competitive process where miners race to solve a cryptographic puzzle (finding a valid hash) that lets them add a new block to the chain. The chance that any single miner is the one to find a block depends directly on its hash rate relative to the total hash rate of the network.
Over the last few years, the Bitcoin network’s compute power has climbed to absurd levels (hundreds of exahashes per second, EH/s), which makes mining a block solo harder by the day. Right now the global network runs at roughly 800–900 EH/s (1 exahash = 10^18 hashes per second), holding an average of 10 minutes per block thanks to the difficulty adjustment.
This is the context in which devices like the NerdMiner showed up, the ones people casually call lottery miners. A NerdMiner is a low-cost, low-power micro-miner built mainly for education and fun, letting any enthusiast “try their luck” at solo Bitcoin mining.
Check out my complete guide to the NerdMiner.
These boards don’t use dedicated ASIC chips; they run on general-purpose microcontrollers. The latest version, the NerdMiner V2, is built on an ESP32-S3 microcontroller (dual-core LX7 at 240 MHz) with a small TFT screen to show stats. The hardware packs 16 MB of flash, 8 MB of PSRAM, WiFi, and runs off USB (5 V) at a typical draw of around 1 W. Despite that tiny power budget, its hash rate sits in the tens of kilo-hashes per second (kH/s).
In practice, a current-generation NerdMiner hits roughly 50–80 kH/s, a figure millions of times smaller than what a modern ASIC does. Figure 1 shows a NerdMiner V2 running, with a hash rate of about 55 kH/s on its screen.
Figure 1. NerdMiner V2 device running in solo mining mode, with a built-in screen showing hashing stats (rate ~55.34 kH/s, no valid blocks found yet).
The reason to run a NerdMiner isn’t profit, it’s to learn and tinker. It does technically take part in the Bitcoin network, but its odds of success are practically zero. Even so, it’s genuinely interesting to put a number on that probability and compare it against more powerful mining gear.
Better than buying a NerdMiner is building one yourself. I wrote this step-by-step tutorial for building a NerdMiner V2. A project that teaches you a ton about mining, programming, and networking.
The method behind the calculation
To estimate the chance of mining a block with a NerdMiner, I use a probabilistic model based on the share of hash power the device contributes versus the whole network. Put simply: if a miner contributes a fraction f of the global hash rate, its probability of being the one to mine any given block is roughly f. A device with 1% of the network’s total power would have a ~0.01 chance (1 in 100) of solving a block on a given attempt. In our case, f is going to be ludicrously small (somewhere around 10^-15 to 10^-16) because the NerdMiner works in kH/s while the network works in EH/s.
The network hash rate (in H/s) and the network difficulty are the two parameters that drive the calculation. As of March 2025, Bitcoin’s mining difficulty hovers around 1.12×10^14 (112 trillion) and the global hash rate around ~8.3×10^20 H/s (about 830 EH/s). On the device side, I assume a NerdMiner at h_Nerd ≈ 5.5×10^4 H/s (55 kH/s) running the latest firmware. The power fraction this represents is:
That value, f ≈ 6.6×10^−17, that is, ~0.000000000000000066, is the probability that the NerdMiner finds one specific block (say, the network’s next block). To get the chance of success over a window of time, I model block occurrences as a Bernoulli process repeated 144 times per day (since ~144 blocks are mined in 24 hours on average) or ~52,560 times per year (144 × 365). Because f is so tiny, the probability of hitting at least one block in an interval can be approximated as P ≈ 1−(1−f)^N ≈ N·f for N attempts (assuming independence and Nf ≪ 1). So the approximate odds of success come out to:
- Per block: P_block ≈ f ≈ 6.6×10^−17. That’s roughly 1 in 1.5×10^16 attempts (more than ten quadrillion).
- Per day: P_day ≈ 144×f ∼ 9.5×10^−15. In other words, ~1 in 1.05×10^14.
- Per year: P_year ≈ 52,560×f ∼ 3.5×10^−12, i.e. ~1 in 2.8×10^11. As a percentage, that’s 3.5×10^-10 %, practically zero.
The flip side of that is the average waiting time to find a block with a NerdMiner. The inverse of f tells you how many blocks, on average, you’d have to try before hitting once. With f ≈ 6.6×10^−17, the average number of blocks you’d need is ~1.5×10^16. At ~144 blocks/day, that works out to ~1.05×10^14 days, which is about 2×10^11 years (200 billion years). That number is astronomically larger than the current age of the universe (~1.3×10^10 years).
One caveat worth stating plainly: these probabilities assume the network’s difficulty and hash rate stay constant. In reality, difficulty keeps adjusting and usually rising, which makes the task even less likely over time.
To put the results in context, I also pulled data on other mining devices. Specifically, commercial ASIC miners across several generations: the old Bitmain Antminer S9 (≈13 TH/s), a mid-generation ASIC like the Antminer S19 XP (≈140 TH/s, released in 2022), and a current-gen one like the Antminer S21 Ultra from 2024 (200–335 TH/s depending on the model). For each, I worked out its power fraction f relative to the network and the equivalent odds of mining a block. I also gathered specs on power draw (watts) and efficiency (energy per hash) to judge the economics. I assume a typical electricity cost of 0.10 USD/kWh to estimate daily and yearly operating costs for the higher-power devices, weighed against potential income from block rewards (3.125 BTC per block mined in 2025, after the 2024 halving).
In short, the method pairs a probabilistic analysis (a Bernoulli/Poisson model for rare events) with real hardware and market data, leaning on reliable, up-to-date sources for hash rates, difficulty, technical specs, and solo-mining experiences.
The results: what are the 2025 odds of a NerdMiner mining a Bitcoin block?
Run the numbers and the answer is blunt: the probability that a NerdMiner (~55–78 kH/s) mines a Bitcoin block is practically zero.
Per individual block, the chance is on the order of 10^-17 (tens of quadrillions to one). In terms of expected time, you’d need ~2×10^11 years of continuous operation on average to land a single block under current conditions.
For perspective, the age of the universe is estimated at roughly 1.3×10¹⁰ years (about 13 billion years). Which means the expected time for a NerdMiner to mine a Bitcoin block is more than 14 times the current age of the universe.
With this probability in hand you’ll get more out of the conclusions in this piece: Is a NerdMiner worth buying?
Figure 2 drives home just how wide that probability gap is, comparing a NerdMiner’s hash contribution against the network total.
Figure 2. Note that the data is on a logarithmic scale.
For comparison, Table 1 lists different mining platforms and estimates their odds of solo success:
| Device | Hash Rate (TH/s) | Probability per block | Daily probability | Annual probability / Average time |
|---|---|---|---|---|
| NerdMiner | 0.000055 | ~6.6×10⁻¹⁷ (1 in 1.5×10¹⁶) | – | ~3.5×10⁻¹² (1 in 2.8×10¹¹) |
| Bitaxe mini-ASIC | ~3 | ~3.6×10⁻¹² (1 in 2.8×10¹¹) | ~1 in 1.2×10⁶ | Block every ~3,500 years |
| Antminer S9 | 13 | ~1×10⁻¹¹ (1 in 1×10¹¹) | ~1 in 8.6×10⁶ | ~1 in 60,000 (~60,000 years) |
| Antminer S19 XP | 140 | ~1.7×10⁻¹⁰ (1 in ~6×10⁹) | ~1 in 48,000 | ~1 in 133 (~133 years) |
| Antminer S21 Ultra | 200 | ~2.9×10⁻¹⁰ (1 in 3.48×10⁶) | ~1 in 24,000 | ~1 in 66 (~66 years) |
| Antminer S21 Ultra (Hydro-cooled) | 335 | Improves to ~1 in 2.08×10⁶ | – | ~1 in 40 (~40 years) |
Table 1. Comparison of hash rates and solo-mining probabilities (approximate) for different devices, assuming ~830 EH/s of global hash power. Even with the latest ASIC hardware (S21 Ultra), the chance of discovering a block solo stays extremely low (on the order of 10^-7 to 10^-6 per day). A micro-miner like the NerdMiner is tens of millions of times less powerful, which leaves its odds of success at practically zero.
The takeaway is a near-linear relationship between hash power and the probability of mining blocks: more hash rate raises your odds proportionally, but even jumps of several orders of magnitude still leave the probability vanishingly small.
Against all odds, one solo miner actually mined a block
It’s worth flagging a real-world case: a solo miner with ~3 TH/s (similar to a Bitaxe) who, against all odds, mined block #853742 in July 2024. According to the solo pool operators, a hash rate like that should find a block roughly once every ~3,500 years on average, or put another way, it had about 1 chance in 1.2 million of pulling it off on any given day. It confirms that, however overwhelmingly low the theoretical odds, the process is random and extremely improbable events can and do happen (just like someone eventually wins the lottery).
Per hash, the NerdMiner burns more than any useful miner
Because it runs on a generic microcontroller, the NerdMiner’s compute efficiency per hash is dreadful next to a dedicated ASIC. It draws about 0.7–1.5 W for ~50–80 kH/s. That works out to roughly ~12,000–20,000 J per trillion hashes (J/TH). A modern ASIC like the Antminer S19 XP, by contrast, draws ~3,010 W for 140 TH/s, landing an efficiency of ~21.5 J/TH.
Which means a NerdMiner burns on the order of 600 times more energy per hash than an S19 XP.
Even the old Antminer S9 (≈13 TH/s, 1,300 W) was two orders of magnitude more efficient (~98 J/TH) than the NerdMiner. The reason is that ASICs use hardware highly optimized for the SHA-256 algorithm, while the NerdMiner computes its hashes in software on a general-purpose embedded CPU.
In absolute terms, 1 W is negligible next to the ~3,000 W a typical high-performance ASIC miner pulls. Over a full year, the NerdMiner would use ~8.8 kWh (kilowatt-hours), costing ~$1 if electricity runs $0.10/kWh. A single Antminer S19 XP would burn ~26,400 kWh in that same period (~$2,640 a year in electricity). The catch is that the S19 XP’s expected Bitcoin output is also billions of times higher.
In fact, none of these solo miners guarantees regular income, which is why the vast majority of participants join mining pools, sharing rewards in proportion to their hash contribution to get frequent payouts. A low-power device like the NerdMiner isn’t even practical for pool mining, since its contribution is so small it would rarely manage to validate a useful share of the block.
If the odds are basically zero, why buy a NerdMiner?
Given the near-zero odds, mining with a NerdMiner should be seen only as an educational hobby. The hardware cost (roughly 40–60 EUR) is unlikely to ever be recouped.
The expected value of the reward is essentially zero. Even assuming a high Bitcoin price (say 100,000 USD per BTC), the expected annual gain would be $100,000 × 3.125 × 3.5×10^-12 ≈ $1.09×10^-6, less than a hundredth of a cent per year on average. That’s far below the annual electricity cost (~$1–2) and the device’s depreciation. In other words, you lose money running a NerdMiner, even if only a little, the same way someone buying a lottery ticket expects to lose the value of the ticket on average.
The educational payoff and the message it carries
Running one of these little gadgets lets you grasp firsthand how it connects to the network, how mining gets configured (entering the wallet address, connecting over WiFi to a node or pool, and so on), how to monitor difficulty, and watch the screen tick along without ever finding a valid block.
It’s a tangible reminder of the huge gulf between a home miner and the combined might of the global network. Projects like this democratize access to mining in the sense that anyone can take part symbolically with a minimal investment, even if they’re practically doomed never to solve a block. Some enthusiasts argue that letting thousands of people run small home miners adds to the network’s decentralization and diversity, even if their real influence on network security is imperceptible.
The psychology of the lottery
One genuinely interesting angle is the psychological, playful side: like playing the lottery, having a NerdMiner switched on can give a little thrill or a remote hope (“what if today’s my lucky day and the block comes up?”). The potential reward is substantial (3.125 BTC per block mined today, worth several tens of thousands of euros), so the appeal of “dreaming” about that prize is real, even when you rationally know the expectation is practically zero.
Bitcoin’s history has its share of extraordinary cases of small solo miners hitting it big (like the 3 TH/s block mentioned above, plus others reported here and there in solo-mining pools), which feeds the “it’s not impossible” narrative. These anecdotes, statistically expected given the sheer number of independent attempts worldwide, keep the hobbyists inspired.
It only gets harder from here
One last point: this analysis is based on 2024–2025 data. The specific numbers (global hash rate, difficulty, BTC price, and the rest) will shift over time, but the qualitative conclusion won’t really budge: mining Bitcoin solo with very-low-power devices will keep being, in probability terms, a near-impossible event.
Short of radical changes to the protocol (a drastic, permanent drop in difficulty, which hasn’t happened at scale) or innovations that turn ordinary devices into competitive miners (which would defy the physics of where mining is headed), the situation stays exactly as it is.
Frequently asked questions
What are the odds of a NerdMiner mining a Bitcoin block?
Practically zero. Per block the chance is around 6.6×10⁻¹⁷ (about 1 in 1.5×10¹⁶), and per year roughly 3.5×10⁻¹² (about 1 in 2.8×10¹¹). On average you’d need around 200 billion years of continuous mining to land a single block under current conditions.
How is a NerdMiner’s probability of finding a block calculated?
You take the device’s hash rate as a fraction f of the total network hash rate. That fraction is the probability of mining any given block. To get the odds over a day or a year, you multiply f by the number of blocks in that window (~144 per day, ~52,560 per year), since the chance of at least one success approximates N·f when f is tiny.
Has any solo miner ever actually mined a block?
Yes. A solo miner with about 3 TH/s mined block #853742 in July 2024, despite odds of roughly 1 in 1.2 million on any given day. It’s the lottery effect: with enough independent attempts worldwide, an extremely improbable event eventually lands for someone.
Is a NerdMiner profitable?
No. The expected value is essentially zero. Even at a high Bitcoin price, the expected annual gain is a fraction of a cent, far below the ~$1–2 of yearly electricity and the device’s depreciation. You should treat it as an educational hobby, not an investment.
Why does a NerdMiner use so much energy per hash?
Because it hashes in software on a general-purpose microcontroller instead of a chip built for SHA-256. Its efficiency is around 12,000–20,000 J/TH versus ~21.5 J/TH for an Antminer S19 XP, roughly 600 times worse. In absolute terms it only draws about 1 W, but per hash it’s wildly inefficient.
Do the odds get better over time?
No, they get worse. Network difficulty keeps adjusting and usually rising as more hash power joins, which lowers any single small device’s share. Barring a drastic, permanent drop in difficulty, the NerdMiner’s odds only shrink over time.
Learn more about the NerdMiner
