bit_by_bit submitted by
publishes a daily mining-cost-per-coin
Though his work is thorough and commendable, it is unfortunately incorrect, and his conclusions naive.
I'm sure he has misled people on this board, so I'm here to set the record straight.
Roughly using bit_by_bit's assumptions:
achieves these costs per coin:
|Miner ||Cost per coin |
|Cointerra TerraMiner IV ||$1985 |
|KnC Neptune ||$2030 |
|Bitfury BF3500 ||$1650 |
|Bitmain Aintminer s3 ||$866 |
|Bitmain Aintminer s2 ||$1305 |
|Average ||$1570 |
- difficulty increase is (probably) impossible to predict
- I'm 100% sure your miner would not arrive and be switched on today
Given those two huge, highly variable (and unpredictable) factors, trying to work out a cost-per-coin is ... more-or-less impossible.
It's simply enough to assume that mining is extremely unprofitable at the moment and (probably) a very poor investment.
Here are some examples of variability:
|Cointerra TerraMiner IV ||cost per coin |
|0% difficulty increase ||$759 |
|20% difficulty increase ||$2640 |
|15% difficulty, but starting at 30bn difficulty ||$3265 |
How are these numbers so different from bit_by_bits?
His calculations do not factor in an exponential difficulty increase. Instead, he says (in his maths) : "if the bitcoin network were composed of the miners here, and no extra miners are added/removed (i.e. difficulty remains the same) what would those miners (on average) achieve as a cost per coin over six months."
The problem with these numbers is
- The percentage of miners he uses to compose the network is unknowable, and as you see above, miner performance varies greatly. I'm quite sure that huge operations custom manufacture their machines and never sell them. Their performance is unknown. (an unknowable unknown)
- The makeup of the mining network in the future is unknowable, and difficulty will undoubtedly increase, but we can't know by how much. It has previously plateaued. Will it do the same? nobody knows.
- They assume the very latest miners, shipped immediately. Historically, new miners are not shipped on time. It's been suggested that the manufacturers keep them and do highly profitable day-zero mining with them.
Also, to suggest that it is possible to predict market movements (and depth) is naive as it asserts that demand is constant, and that supply is the major, or key, factor. This is highly unlikely to be the case.
Let's talk about mining's effect on bitcoin price or, first should we talk about the effect of the price of bitcoin on the mining industry?
The two are intimately linked chicken-and-egg in a feedback loop.
For a manufacturer to decide to make a rig, they need to design chips, get industry contacts, produce things (in china), make sure they work, then ship. They also need to get orders and decide if they are able to get the whole project in time for market.
These projects are multi-month/year, and I've heard success is largely decided by who you know
in china (china's pretty busy already).
There is some kind of lag. Investors also pre-order, and must take a wild guess at future conditions with no guarantee whatsoever.
At times like now, where mining is so unprofitable, which miners are actually selling coins (at a loss)?
Large operations have large overheads, but to sell now, when the price might rise by 10x again would be idiotic.
So, really this "supply" aspect of the supply-demand equation is very difficult to get a decent hold on, though I would love somebody to attempt it as a PHD
. The blockchain should provide some answers.
The other side of it (what miners will be produced) is also difficult to know. It could be that right now (with an unprofitable industry, and miners actually being quite close to desktop PC chip-size - i.e. as fast as humans can make them) no miners are in the pipe-line. This could (in crazy theory) lead to a zero difficulty increase for the lucky new owners of the above rigs. In that case, bit_by_bit's numbers would be spot on.
Unfortunately, it's absolutely unknowable.
So... why do people buy miners now?
Quite simply, getting your head around an exponential anything
is hard. The exponential difficulty increase is a motherfucker. But it's good for bitcoin (it protects our network from meddlers).
Also, you could gamble that mining difficulty has
to slow down... surely...
In my experience, looking at price charts is far more informative about future market movements. But, whilst I've got the microphone, I would remind newbies not to trade their coins
Show your workings!
Using this mining profitability calculator
and inputing the "profitability decline per year" from this tool
I'll work through Cointerra TerraMiner IV as an example.
|meh ||meh |
|Difficulty ||18,736,441,558 |
|ghash ||2000 |
|Elec ||$0.15 (varies quite a bit from country-to country, like 0.7 canada to 0.2 UK?) |
|power ||2200 |
|time frame ||6 months |
|hardware cost ||$6000 |
|price per BTC ||600 |
|profitability decline ||0.00882406 |
I got these numbers off bit_by_bit. I don't care about the details. My argument is that it's not an answerable question.
|meh ||meh |
|Electricity cost ||$1446 |
|Total cost ||$7446 |
|Income ||2553 |
|Coins ||2252/600 = 3.75 BTC |
|$/BTC ||7446/3.75 = $1985 |
Please, if I've made a mistake, let me know and I'll send bit_by_bit some flowers.
"Why are you just posting stuff directly against another user: that's not cool"
Well, it's whatever motivates you eh?
I just go wound up by our discussions. But, I'm quite sure there are people on this board who don't know this stuff, so ... it's probably beneficial.
Have fun EDIT: Ok, so I genuinely thought that I had made a fact-based post. Er, I added a few comments that I thought were funny, but I guess that wasn't a great idea. I removed one of my comments myself, but it's true that the moderators were in touch.....
And - to bit_by_bit, I am sorry, because some of the things I said were above and beyond "spirited discussion".
I absolutely agree that polite conduct is the way forward, and my initial "hang on a minute" reply to him was nice.
But, I do have to admit that this subject has wound me up a fair amount. I genuinely believe that he's made a quite serious mistake - but I am happy to be proved wrong. Right now - I just want to get to the bottom of this.
I am a miner I didn't want to add this before, because I'm sure it (incorrectly) gives my argument more weight. But I need you to understand that bitcoin difficulty is a total motherfucker.
I pre-ordered a BFL single for 11BTC in May 2013. The difficulty was about 4 million, and I worked out I'd make 30BTC/day at those conditions.
It arrived at around 30 million difficulty, and I think now we're 18 billion.
I've made about 0.7 BTC mining, and It's on the limit of believability that I'll make 1BTC before I throw it in the bin. I have a suspicion that it will be useful in the future for some altcoin/blockchain like thing. Also, I got free heating (which was the whole reason I discovered bitcoin in the first place!)
Horrific loss. I think it makes about $1 more than it costs in electricity to run (at current price......)
This whole post is not a "bitter miner" but somebody who has experienced bitcoin's exponential difficulty First Hand. Honestly, it is unbelievable.
I genuinely think that the guy that does the profitability calculator deliberately does not explain what the 'profitability decline per year' is ... because he knows it will adversely harm bitcoin and the manufacture of miners.
Even More EDIT:
Am I sure I've got the difficulty increase thing right? So, I've made a spreadsheet thing to see if the 0.0022 difficulty thing is right. It is.
All this table tells you is that in order to calculate 15% difficulty increase, you need to use a number LIKE 0.0022 in the 'profitability decline per year' box, and not 0.98 (which bit_by_bit calculated).
I've sanity checked my numbers against the 'profitability calculator' and they don't quite line up, but they're close enough. The difficulty is not the same either, but it's in the same region. I don't know why. Also, the months aren't exact fortnights, so they don't line up. These are details.
This proves my above workings to my satisfaction.
what is this horrible data? It shows how much BTC your miner earns each 2 weeks (average difficulty change period). The last 2 rows (calc:) are from the profitability calculator website (and so are right). My attempt is on the left. Fortnight 13 is 6 months.
Oh, this graph uses 14.07% difficulty.
2 bold numbers. 1 is approximately the 3.75 coins that gives you $1900 / coin whatever. 2 is the "profitability decline per year" as a tiny number. The pro tool comes up with 0.01095125 and I got 0.02257 but I don't care - it's close enough.
|fortnight number ||"profitability decline" ||BTC earned ||BTC earned accumulated ||calculator says BTC |
|1 ||1 ||0.7518 ||0.7518 ||x |
|2 ||0.8593 ||0.64602174 ||1.39782174 ||1.361666667 |
|3 ||0.73839649 ||0.555126481 ||1.952948221 ||x |
|4 ||0.634504104 ||0.477020185 ||2.429968406 ||2.295 |
|5 ||0.545229376 ||0.409903445 ||2.839871852 ||x |
|6 ||0.468515603 ||0.35223003 ||3.192101882 ||2.938333333 |
|7 ||0.402595458 ||0.302671265 ||3.494773147 ||x |
|8 ||0.345950277 ||0.260085418 ||3.754858565 ||3.378333333 |
|9 ||0.297275073 ||0.2234914 ||3.978349965 ||x |
|10 ||0.25544847 ||0.19204616 ||4.170396125 ||3.68 |
|11 ||0.21950687 ||0.165025265 ||4.33542139 ||x |
|12 ||0.188622254 ||0.14180621 ||4.477227601 ||3.888333333 |
|13 ||0.162083103 ||0.121854077 ||4.599081677 ||x |
|14 ||0.13927801 ||0.104709208 ||4.703790885 ||4.03 |
|15 ||0.119681594 ||0.089976622 ||4.793767508 ||x |
|16 ||0.102842394 ||0.077316912 ||4.871084419 ||3.213333333 |
|17 ||0.088372469 ||0.066438422 ||4.937522842 ||x |
|18 ||0.075938463 ||0.057090536 ||4.994613378 ||4.195 |
|19 ||0.065253921 ||0.049057898 ||5.043671276 ||x |
|20 ||0.056072694 ||0.042155452 ||5.085826727 ||4.241666667 |
|21 ||0.048183266 ||0.03622418 ||5.122050907 ||x |
|22 ||0.041403881 ||0.031127437 ||5.153178344 ||4.273333333 |
|23 ||0.035578355 ||0.026747807 ||5.179926151 ||x |
|24 ||0.03057248 ||0.022984391 ||5.202910542 ||4.295 |
|25 ||0.026270932 ||0.019750487 ||5.222661028 ||x |
|26 ||0.022574612 ||0.016971593 ||5.239632622 ||x |
My whole point is that these numbers are totally unworkably all over the place. You can't calculate them meaningfully.
Dear Bit_by_Bit I CANNOT BELIEVE the amount of effort that I have had to go to in order to show you that you made a minor mistake. (at time of writing you still deny it).
There is no doubt in my mind now that I was right in the first place. Your calculations do not include a significant difficulty increase.
I wish you well.
First and fore most; Yes I am new to this board. No I am not new to the ideas of reddit, mining or of a coop. submitted by
I was wondering if the members of reddit would be interested in forming a Coop. I just recently started one for another group of friends (no worries, no links or the like here) and it occurred to me to see about one for reddit, given its large user base and number of enthusiasts. What I did before was set up a site with a webstore applet (and a twitter feed, about us sections, rules section, etc) linked to paypal so that a legal chain of custody for the funding was established and recorded (bitcoin transfers being anonymous and all). This also established publicly on the site a set of terms and conditions for the Coop. What we came up with for ground rules were as follows;
Payout % will be based on % of hashing power purchased (in other words if you buy 1/100 of the machines power for yourself, you get 1/100 of the net pay out)
Payouts will take place on the 15th and 30th of each month after mining becomes active
Payouts will be in BTC transfer only, and all members are required to have a Bitcoin wallet address
Hashing power will be sold in blocks of 50 GH/s for the first machine (40 blocks in total)
A 3% upkeep and administration fee will be deducted before payouts take place
25% of total revenue will be reinvested for the purpose of increasing our network power. Members will be given equal stake in new equipment equal to there previous % at no charge.
The network will purchase Terraminer IV machines until/unless something with a better powecost ratio is on the market (these machines currently cost $3/GH before shipping and taxes, approximately $6,800 with shipping and taxes included)
The coop will never forcibly buy out anyone's stake in the network
A network member may request a buyout equal to 4x a normal payout. However they may not do so for the first 6 months of active mining
Mining will begin in April and a twitter account will be set up on behalf of the Coop by its manager so that he can keep everyone updated in a timely manner
Once all the stakes are bought, no new members may "buy in"
Members may not sell their share of the network to any one user to prevent general BS and ill feelings of back alley dealings
If the network should ever receive a request for buy out and do so, that share of the network is to be evenly split among remaining members. Members will not be charged for this and the reinvestment % will be used to conduct all transactions of this nature
So there you have it. These rules can be changed to suit need if anyone has ideas about them or additions they think necessary (just make sure your idea is concise and clear).
Thoughts? Opinions? Feedback?
Der Terraminer IV von Cointerra war wohl kein Erfolgsgarant. (Bild: Cointerra) Das Unternehmen stellte nicht nur ASIC-basierte Mining-Rechner her und verkaufte sie, sondern war auch im Mai ... CoinTerra cuts price of TerraMiner IV bitcoin mining rig CoinTerra has just announced its official launch and, with this, is cutting the price of its TerraMiner IV product from $15,750 to $13,999. Cointerra Terraminer IV Review: Cream of the Crop, for NOW Dedicated Bitcoin mining computers, aka ASIC (Application-Specific Integrated Circuit) Cointerra IV Setup for Bitcoin Mining The Cointerra IV was launched for public purchase in January of 2014. CoinTerra cuts price of TerraMiner IV bitcoin mining rig David Gilson Aug 27, 2013 CoinTerra has officially launched and announced a price cut on its main 28nm bitcoin miner. Currently the company sells their TerraMiner IV 1.6 THS Bitcoin Miner for $3499 USD or less than two times cheaper compared to the 2 THS cloud mining plan they have available for 1 year priced at $7999 USD. And there is another thing that is not very clearly defined is if there will be a maintenance fee for the cloud mining hashrate you are going to be purchasing. There is a brief mention of 5 ...
This video is unavailable. Watch Queue Queue TerraMiner IV by CoinTerra Performance Follow-up Video (24 hours) - 1.5TH/s - Duration: 2:56. ... News Bits on: Bitcoin in New York - Duration: 2:34. Bitcoinist.net 678 views. 2:34 . Bitcoin ... Welcome back truth seeker! It's me, Zachary K. Hubbard, and this is the 20th channel due to Google's censorship. If you're new here, this channel is dedicate... This is the unboxing video of the new TerraMiner IV from Cointerra. In this video we unbox, do a quick run through. In the next video we will show it running and the specs. Here is the follow-up video promised in the unboxing video. We let it run for 24 hours exactly. We are getting 1.5TH/s on average and it has drifted higher. Next video we will hook-up two power ...