More than anything else this is probably determined by electricity cost varying in different areas.
In my area we pay $0.10/kwh but anyone paying above approx. $0.17 has either shutdown or is thinking hard about it now. The variance of cost is wide - from "free" in shared dorms, to 0.01 in Venezuela to 0.40 in some European countries. Even the tech haven of California has pretty nasty electric rates.
I wouldn't be surprised if someone put together a map of miners from over time that we'd see whole areas go dark over time like a grid shutting down. As the price drops miners are forced to confront whether they want to spend electricity to hold bitcoins for future possible price increases, or just go dark.
And in sync with this the price of 5830 gpus on eBay has dropped from rough $135 a couple months back to more typically $100 or less now. This is all a natural outcome of people getting into mining to make money rather than for philosophical reasons, though there is always some blurring of the boundaries.
Keep in mind that as difficulty falls this improves the return and it should slow the hash rate decline. So everyone is watching that too and deciding what they consider the break even point to be.
Topic: Slow loss off network power (Read 1232 times)
why worry? i'm happy as difficulty falls. pay day is when I'm out of fiat money and I hope this is many years from now with bitcoin really growing strong.
Exactly. Some of the pools provide both # of users and total hashing power. This lets us get a guestimate for hashing power per capita. It looks to be around 400MH/s per capita (well technically per account). So 1 miner dropping off is only a 0.0004TH change. You can lose hundreds of miners a day and it will still appear "smooth" and gradual when looking at the entire network.
We have lost ~2TH from the peak so that is thousands of miners individually idling rigs when at different times over the last month.
We have lost ~2TH from the peak so that is thousands of miners individually idling rigs when at different times over the last month.
So what you are saying is that a lot of people have stopped mining. If each person is a small amount it must take a lot of those small amounts to equal the drop in hash power.
Each miner decides to quit at a different time. Some decide to idle only some of their rigs (least efficient ones).
It simply looks "smooth" because you are looking at it on a TH scale. Each individual rig going dark is a tiny amount creating the perception of this slow consistent decline.
It simply looks "smooth" because you are looking at it on a TH scale. Each individual rig going dark is a tiny amount creating the perception of this slow consistent decline.
As the market price falls, those miners who are least efficient will turn off their machines. As marginal profitability drops, so too does hashrate. When price rises, those miners at the margins will turn machines on again. It's both a normal and expected mechanism within the protocol.
So, it seems like we're losing hash power at a steady rate. How is this possible? It seems too steady to be a natural occurrence. I would have expected large drops with some steady states and not this slow trickle...
Jump to: