According to data from Blockchain.info, Hash-rate dropped yesterday from around 100 exa-hash per second all the way down to around 68 exa-hash per second, a 30% drop. While there is no obvious explanation for the drop, we can speculate the cause. From our experience in the industry and its current state, the most likely reason is one or more large miners upgrading their hardware, such as Bitmain moving their operations from S9s to T17s or S17 Pros or any whale miner making a similar move. It is prudent to make this sort of upgrade over the next 8 months, and we may see similar drops as other whales follow suit, because S9-generation hardware will become functionally obsolete at the next Halving.

At the Halving, the S9s profitability will be significantly hindered, likely leading to a price drop and relocation of the machines to the lowest price points for electricity, where they will still maintain some semblance of profitability. New miners are 2–3x more efficient than the S9, and will survive the Halving with their profitability intact. As the S9-generation hardware goes offline we can expect a large drop in difficulty following the drop in hash-rate. We foresee two scenarios: one, the price of Bitcoin rapidly rises before the Halving. In this scenario, S9-generation hardware will maintain its profitability and stay online, in which case difficulty and hash-rate will likely both spike as newer generation hardware comes online in other locations. If there is no significant increase in price, we will likely see a plethora of S9s hit the open market at reduced prices, perhaps bottoming out around $100 or even below.

One very important thing to note is that this chart shows estimated hash, based on the block time and the difficulty. It is not technically possible to tell exactly how much hash-rate the Bitcoin network has at any given moment, but it is possible to reverse engineer and estimate hash-rate from the time it takes to solve (Block-time) an equation we know will probably take you X random attempts to solve (Difficulty).

The above chart shows the block time over the past three weeks — note the massive spike on the 23rd of September directly corresponding with the drop in hash-rate. Since mining is all about probability it is normal to see some variation on the block time from block to block, but averaged out over a day or more this variation diminishes and starts demonstrating true changes in mining hash-rate. A group of miners all going offline at once, decreasing the computational power of the network, would explain the reasoning behind such a spike. Less miners working on solving the next block means it will take longer to solve the block.

Regardless of the underlying reason for this large drop in hash-rate one thing is for certain.

With next generation miners such as the S17 Pro or M20S shipping daily, the mining space is in the process of a large-scale, network-wide upgrade, a byproduct of the planned obsolescence integrated into bitcoin’s core code in the form of the network block reward halving.