I would still argue that if the internet breaks down, we'll have more serious problems than what happens to Bitcoin. But I understand that this isn't a satisfying answer. So I'll try to elaborate beyond that.

Hypothetically: no internet, but somehow magically power network is still in some decent shape (unlikely scenario, though).
Bitcoin nodes can't talk via internet with each other, no software updates can be distributed or published. Bitcoin miners are isolated and it doesn't really make sense to work on an isolated chain, because every miner would work on an isolated chain and risking that some other miner's chain could be the heaviest (PoW-wise).

The miner with the biggest hash rate available at one single geographical spot (not needing internet to pool hash power) would be in an advantageous position. They could possibly mine the heaviest valid chain and be sure it will survive when nodes get reconnected again.

All isolated miners would mine more or less "empty" blocks (only with a coinbase transaction) or with some (silly?) own transactions because no other nodes can provide pending transactions to go into a mempool.

In such a hypothetical situation it doesn't make much sense to me to continue to mine as long as nodes can't talk with each other.


From my understanding of Bitcoin (if I'm wrong, I'm more than happy to be corrected) the current protocol has no really working things to cope with a major disruption event that would shatter the current network hashrate.

Current difficulty needs a global hashpower in the ballpark of 820+ EH/s to produce on average one block per 10min (more accurate 2016 blocks per two weeks). Difficulty is only adjusted every 2016 blocks.

Worst case would be a major disruption just after a difficulty adjustment. Assume after a major global disruption event the global hashrate is only 1/10th of current and can't ramp up for whatever reasons: the next difficulty adjustment would take on average 20 weeks, every block wouild be mined on average after 100min.

The next difficulty adjustment can't drop to 1/10th of current, because adjustments are currently limited to max. 1/4th on the low side and 4x on the high side.

If Bitcoin had to deal with some major disruptive event that severely changes global hashpower some node software fork would be required to cope with this quickly.