What was the difference in hashrate power when the March 2018 anti-ASIC Hardfork occurred?

Question

What was the percentage drop in hashrate during the 2018 march anti-ASIC hardfork before and after the fork?

Finding this number, can we estimate how many ASIC's were produced?

This is assuming this drop was all due to ASICs, but I'm sure many were also just due to non-upgraded miners. But judging from this drop, we can get an close approximation to how many were produced.

Answer 1

The fork was moved to today (April 6, 2018), so it is hard to tell. Early indications based on block-times are about 80% of the hash-rate is missing. Some of that is non-ASIC miners failing to update on time. We should know more in a week or so.

Answer 2

The network hash rate was around 1123 MH/s at the time of the fork. The difficulty hasn't changed yet, so blocks are coming in much slower(lower hash rate at same "asic difficulty"), and while it will slowly decrease with each new block, it will take 720 blocks before the true adjustment takes place(a few more days at minimum). As the difficulty is a necessary calculation in overall network hash, we wont really know until after block 1546721 as to where the total network hash will settle.

Answer 3

As of writing this answer (2018-04-07 03:00 UTC), ~18.5 hours after network upgrade, there are 65 blocks mined (latest is 1546065). It takes on average 17 minutes to mine a block, which is 8.5 times the expected 2 minutes.

The difficulty is calculated based on past 720 blocks. However not all blocks are used.

DiffCut: the number of highest and lowest timestamp values to be ignored, as they are considered to be outliers. Default value: 60.

DiffLag: the number of last blocks that should be discarded previous to any subsequent computation. This is done to make it harder to create a blockchain fork with higher cumulative difficulty. Default value: 15.

When there is a sudden drop in hashrate, the time taken to find blocks will increase. However, the latest 15 blocks are always ignored, and then the 60 blocks with longest time are ignored. (The 60 blocks with shortest time are also ignored, but we are not interested in them in this calculation.) The difficulty will only start to drop beginning from 1546075. At 1546065 which is before 1546075, we are still working at pre-network-upgrade difficulty. With block time being 8.5x the original, we can estimate the current hash rate to be 2/17 of original.

The reported "network hash rate", which is solely based on difficulty, is still 1145 MH/s. This number actually refer to the real hash rate before network upgrade. (Recall that we are still before 1546075) Using this, the real hash rate after network upgrade is estimated to be 134 MH/s.

The hash rate loss is about 1010 MH/s. If we put all blames on AntMiner X3 (220 KH/s, IIRC miner with highest claimed hash rate), a simple division shows that there are about 4600 units. Similar maths can be done using other ASICs.

Note:

1. These calculations carry large error due to the small no. of blocks (65) used, and within these blocks the block time vary largely. This Reddit thread gives more details.
2. The decrease in hash rate is not solely due to removal of ASICs. Botnets, legit miners, and mining pools that failed to upgrade will also contribute to the hash rate loss. Later on when everything other than ASICs have upgraded, and once beyond 1546720 when difficulty is calculated without the effect of ASIC, we can get a more accurate number for ASIC-only hash rate. (or not, because of the influx of miners from other coins since Monero mining is profitable again)