An old post from a Bitcoin forum, published on September 7, 2010, has recently drawn attention again. The poster estimated the power consumption, computing power, and electricity prices of the equipment at that time, and calculated that if mining alone, the expected cost of electricity to obtain a reward for one 50 BTC block would be about 5.68 US dollars. However, he still referred to this expense as a “loss.”
This set of numbers does not mean that spending $5.68 will definitely result in the mining of 50 Bitcoins. It is merely an estimated power cost based on the network difficulty at that time, individual computing power, and average waiting time, used to measure the approximate economic viability of solo mining.
338 hours is just the average waiting time.
This forum user with the username TTBit claims that their computer consumes about 140 watts of power and has a computing power of approximately 2,200 kilohashes per second. According to the mining calculators at that time, it would take an average of 338.05 hours, which is about 14 days, to have a chance of finding a valid block.
Calculated based on this time, the total power consumption of the device is approximately 47.33 kilowatt-hours. If the electricity price is $0.12 per kilowatt-hour, the total cost of electricity is about $5.68. There are no obvious issues with the arithmetic in the original post itself.
But this is only an average in a probabilistic sense. Miners may produce a block quickly, or they may wait longer, or they may not be able to produce a block at all before the difficulty changes. Therefore, it is not accurate to interpret $5.68 as a “fixed production cost of 50 BTC”.
At that time, the value of 50 BTC in US dollars was not high.
In 2010, the block subsidy for Bitcoin was 50 BTC, but it has now dropped to 3.125 BTC. The protocol stipulates that the subsidy is halved every 210,000 blocks, and the most recent halving occurred in April 2024.
The problem is that in 2010, the Bitcoin market was still very immature; transactions were mainly spread across small platforms and over-the-counter transfers, and price records were not standardized. Public data generally shows that Bitcoin was below $1 throughout most of 2010, with common quotes towards the end of the year ranging between $0.10 and $0.30.
In such an environment, even if one can obtain 50 BTC on paper, it may not be easy to sell them, let alone realize a stable profit. The original poster considered it a loss, which is more in line with their personal judgment of price, liquidity, and transaction risks at that time, rather than a precise formula for calculating profit.
GPU is set to quickly replace CPU in household use.
This old post also records a technical transition in Bitcoin mining. TTBit mentioned in the post that their own CPU device had only about 2,200 kilohashes per second, while at that time some users claimed that graphics cards could achieve around 25,000 to 33,000 kilohashes per second, with significantly higher energy efficiency.
During the discussion, there was also a claim attributed to Satoshi Nakamoto that a system with 24 cores and AMD could achieve approximately 66,000 kilohashes per second. Although these figures come from exchanges among forum users and are not from unified tests, they do reflect a trend: more efficient hardware is beginning to rapidly widen the gap.
Thereafter, GPU was replaced by FPGA and ASIC. Nowadays, the computing power unit of Bitcoin miners has long since increased from kilohashes to terahashes, and even petahashes, and the overall computing power of the network is incomparable to that of the early days.
The mining competition logic has not changed.
The article points out that although the scale has completely changed, the core issue faced by miners remains the same: whether their expected income can cover the costs of electricity, equipment, cooling, and maintenance.
The Bitcoin network adjusts the difficulty every 2,016 blocks in order to maintain an average block generation time of about 10 minutes. When the computing power increases, the difficulty usually rises; when miners withdraw or block generation slows down, the difficulty is reduced.
As of September 8, 2026, the estimated computing power of the entire network has approached 966 EH. According to the calculations in the text, this is approximately 439 trillion times the speed of that computer from TTBit's time. Even though the difficulty level has decreased in recent years, it has not made mining competitive again; the mining industry is still dominated by large-scale ASIC operators.
This old post that has been brought up again is more like a retrospective look at the reality of Bitcoin in its early days. It shows that, given the price levels, liquidity, and hardware conditions of 2010, it was a reasonable judgment for miners to consider the expected electricity costs of $5.68 to be unprofitable.
