Bitcoin Forum

1. if you have to pay for electricity, you should not be getting into the mining game at this point, period.

2. if you have to pay for electricity, but you already have 1 or several dedicated rigs running, you should still mine, but start winding down your operation, start selling some rigs. You don't want to be selling them at a much lower price when the mass exodus begins (when difficulty reaches equilibrium with electricity cost, those who have to pay for electricity will inevitably start selling their rigs).

The fact of the matter is difficulty will keep rising until it reaches equilibrium with average electricity cost. The price of BTC will keep falling, because the supply of new BTC each day is overwelming the market, unless another wave of significant adoption comes, BTC price will continue to fall slowly. Until block reward is halved sometime in 2012.

I wont lie when I read the subject I was expecting a post from angelus lol.

Your advice is generally sound.  People paying more for electricity (25c/kwh) will get out much sooner than those paying 10c.  But ultimately, every miner will face a choice of continuing to mine with losses (in the hope of BTC's price rising in the future) or giving up.  We've already seen what happens when miners exit in droves from Namecoin.  Difficulty remains high for weeks as few miners can be bothered losing money to keep mining until the next reset.

The BTC market is already very well supplied.  We're creating thousands of new bitcoins per day that must find a buyer today, or a buyer tomorrow.  We could see a mad rush for the exits if people begin to think 'this is as good as it gets'. 

I scaled back a while ago and have only limited mining still churning away.  The hardware was sold off while it was still worth something.  It went for around 80% of new cost, which was pretty good. 

Looking at MtGox bids/asks at the moment, there's a reasonable number of bids, but at around $11.  Most asks are clustered around $13 and $13.50 with a long tail towards the current $11.60 strike price.  We could see some pretty dramatic moves in the next few days.  The only prediction I would make is that we won't be stuck at $11.50 in a couple of days.

I agree people who pay 25c/kwh will definitely see an end to serious profits soon. However people can still make good money off mining as long as they don't live in an area with hugely expensive electricity.

I feel its safe to say any frugal miner can build a machine for 1.5 megahash per dollar. Lets say they invest $1000 dollars today thats 1500 megahash. Lets double the difficulty just for the hell of it. I doubt it will get to 3,780,724 any time soon especially if the price drops any more but I feel like no one will argue if I am more then fair. At that difficulty you would make .4 btc.

Lets say the price stays around 12 dollars (if the price drops more I highly doubt even more then this I highly doubt we will reach 4 mill diff) thats 4.8 dollars a day before electricity.

1500 megahash would probably require about 1kw. If you don't agree I measured my watts used for 4 5830's and its ~888 and that gives me 1320.

At 1 kw and 10c/kwh it would cost 2.4 a day. That means 2.4 dollars a day in profit which is 72 a month. 72 dollars a month may not sound too great but thats months down the road and by then any hardware bought now should be easily paid for.

Must be Anusgelus' other account.  I didn't see these posts yesterday.  Must be a sign to start buying

It isn't clear at all that BTC price will reach equilibrium with electricity cost. It could do that for a short time, but in order for it to do that long term, I think it would have to be that case that everyone who continues to mine is just an enthusiast mining for fun. While there are a lot of enthusiasts out there, I don't think they would mine on a large scale. And of course if bitcoin completely fails to catch on as a currency, mining could become unprofitable even if the whole network drops to a few hundred Mhps. It may be that now is a good time to get out of mining altogether, but zero profit for mining strikes me as a relatively low probablility worst case scenario at this point. The price of BTC can and has been very volatile, so the price could drop to $8 next week, and be back up to $15 a few weeks later. If we do see an extreme price drop, personally I think it would be a good opportunity to pick up cheap GPUs or mining rigs from miners who get overly emotional about the situation and bail out.

How can you possibly get 1500 MH/s for $1,000?  I would love to see where to buy this setup.  In fact, I will take 5 completed systems if you can do it at that price. 

uh.

5850x4, overclock like there's no tomorrow.
buy everything used.

costs me like 875.
still gotta wait until all of the parts arrive though.  :-\

5850s:

These can do 400 MHashes each, so you need 4 @ 145 = $580 I personally bought some last week at this price.  you just need to look for them, they become available sometimes.   (you could also do 5 5830s if you find them under $110)

Can you fit the rest of components into $1000 - $580 = $420 (!) ?  I bet you could without breaking a sweat.

There you have it.  Go build your 5 systems.

kokojie thanks for the warning !

So let me get this right:

step 1)  buy 4 video cards that are not readily available, and the ones that are easily exceed your claimed "price"  (at least 200 a piece)
step 2)  buy 1250 Watt PSU, 4 PCIx MOBO, processor, RAM under $420   ???
Step 3)  profit!

Foolproof.  

There's a ton of ways you could achieve this but I'm too lazy to look up specific numbers. People above me have the right idea.

I want to start by reposting what I said earlier.
Next I want to say the 5850 consumes a lot less power then a 5830. My 4x5830 set up consumes 888 watts so a 1000 watts for 4 5850's isn't out of the question. A 4 pcie mobo can be picked up for less then 100 dollars easily and a cpu only costs 30-40 ram is 10 so its extremely do able.

I should have prefaced with this:  I DON'T WANT SOMEONES USED CRAP.

Show me where I can build a NEW system that gets 1500 MH/s for $1,000. 

Wrd, the 5850s crap all over the 5830s for MH/s and power consumption. 

3) I seriously doubt that you will ever profit!!!

2) no need for 1250 Watts.  if you are so bent on having 1250 Watts, then get 2 650-750 Watts PSUs ($140 total).

1) I bought 10 5850s last week for $1,492.64 (includes shipping) from ncix.com. you have to look for deals.  alternatively you can use 5 5830s as well (at a good price).

I pay 5.8$ for 250 kw  ;D

similar deal.

i got 2 of the 5850 new for 174. ncixus had a sale....shipping is a bitch though.
mobo, 1x riser, cpu, ram..and/or HD, the total will cost you like 900somethingsomething.

if you want to get the same deal, well, gotta wait until it shows up again.


but then again, i don't care about using someone's used crap, as long as it can pay itself back.

if you have the $$ to waste, go ahead and buy new stuff.

Am I really going to have to list out links because your too much of a dumb ass to realize 420 dollars for non gpu parts can be easily done....
http://www.newegg.com/Product/Product.aspx?Item=N82E16819103944
http://www.newegg.com/Product/Product.aspx?Item=N82E16820134657
http://www.newegg.com/Product/Product.aspx?Item=N82E16820139292
http://www.newegg.com/Product/Product.aspx?Item=N82E16813138324
48 for extenders
http://www.newegg.com/Product/Product.aspx?Item=N82E16817256057

That leaves 78 dollars from the 420 budget....

so you want me to use a MOBO with 2 PCIx slots? 

It's got 4 slots broseph do your homework before you start arguing.

PCI Express 2.0 x16 =2

PCI Express x1=2

lol....."broseph"....awesome. 

But seriously.  I will give you $5,250 to make me 5 of these (must be properly configured for 1,500 MH/s each)....that should leave you a very large profit margin. 

I can make rigs that get 1.2 ghz with a better than 1.5 ratio... does that count? ;)  (all new parts too)

Exactly.. hes posting doomsday on the Discussion section..

You most definitely can get better then a 1.5 ratio but it tends to take a long time to mass produce something like that and you usually have to cut corners like excluding any form of a case.


If you're actually serious pm me and we will try to figure something out.

absolutely counts....can you give me some details?  I need cases.....these go in an office building, and it does not help "our look" to have computer parts strewn about. 

Does this come with an 80+ gold PSU? That is one corner that is terrible to cut. If it does, plz share because i have no idea how you plan to do this. Also, MIR shouldn't be deducted from the price.

If you would have dated this 08/03/2010, it would have been perfect.

It's almost guaranteed not to. I'm guessing it would be entirely caseless and one of the cheapest psu's which isn't ideal at all on the large scale.

I was curious, so I checked out the price of kWh from my utility. Not bad. $0.0950 per kWh. I need to get a kill a watt meter for my pc.

Just because I have time I'll try to build a high ratio build that cuts all corners with the 5850 @ 145

http://www.newegg.com/Product/Product.aspx?Item=N82E16813138324
48 for extenders
30 for cpu (heat sink not included)
10 for heatsink
6-flash drive
10-ram
580 for 4 5850
90 for psu
total=834
expected megahash=1,600

ratio = 1.91
I guess if you want to add a case add ~$60
ratio=1.79

Second.

And with this comment, I will not entertain anymore of your posts until you spend a good day reading as many posts as you can to learn first, then ask good questions later.

What, no one saying the OP is just wanting people to stop mining so the difficulty will go down for him?

We're all quite used to that by now Tiago, we just tend to assume it comes first and foremost from Angelus  ;D

Awesome.  I will consider us even then. 

I'd be more than happy to build you as many systems as you want for similar prices above. Of course, my fee for doing all this is not included in that price, and if you're too lazy to do it yourself, it's not gonna be cheap.

I don't really think he is serious about letting us build rigs for him. If he is though he directed it to me originally and I honestly would take $250 dollars from the $5250 he mentioned because I make theoretical builds anyway on my free time :P

don't build anything for anyone. Just tell us the parts for your unrealistically cheap system.

That's exactly what costs 250kW for ~14min at $0.10/kWh, an average US domestic price.

if you just build your own case with metal from a reputable metal seller (onlinemetals, not the one in the "how to" guide in hardware), it will cost about $20. (assuming you have spare screws, a caliper and a power drill with drill bits)

Yes the rate seems low but that is the rate you pay while you are in the "normal usage" tier. I don't know exactly what they call it. But once you go over the kwh/month "limit" or baseline then you start getting ridiculous charges, up to 14 cents per kwh.

So in other words so long as you are in tier 1 "normal usage" you pay $0.0950 per kwh, tier 2 "excess usage" you pay (hypothetically) $0.5000 per kwh, tier 3 "overboard usage" $0.14 per kwh.

I have the exact same rate. You know what the electricity bill was for my 24/7 dual 5850 miner? $250. Last month's bill, not mining at all? $70. No, I don't use the AC.

Well iv finished mining now, Still holding on to a few BTC which ill sell if price ever gets back to $2X.XX.
My power billing is a bit like that.
Looking on last months bill I got so many KW/H charged at 0.19 then the last day it went up to tire 3 (high usage) which was charged at 0.29 but luckily it only cost $6 for that day but would of sucked being on that pricing for a week or 2. On the low rate it was only costing $2.XX per day about $15-16 per week for my rig and I was making $62NZD a week (when BTC was at $15USD.)

Was a good run tho. Managed to Pay off 1/4 of my 6850, 100% of my 6670 and buy 2X antec-kuhler-h2o-620 watercooler setups.
Both GFX cards are getting a good using with BFBC2 online now.

Just wish I had a time machine tho. Take my rig back in time with the GFX miner client to when bitcoin first started then I would have had sh*t loads of btc to sell when it was at the $2X.XX high.

this is serious business. if you started at that time with such a huge hashrate, it would have affected bitcoins in a very dramatic way. so that's not a good idea.

Yeah, IF you waited that long for BTC to reach $20.  Back when Bitcoin mining first started, even if you had 20 GH/s of hashing power, you still wouldn't be as profitable as you are right now.  I don't get why people are crying about how "unprofitable" mining has become.  Even at $8.70 per BTC, it's still way more profitable than when the earliest adopters were mining.

DELETED for privacy

Where is "here" ?

DELETED for privacy

Switched my 2 mining rigs off today when the price fell to $6.  Will reconsider if it goes above $9.

Thank you!

And this ladies and gentlemen, is a prime example of the Tragedy of the Commons.  A miner quitting as it's pointless mining for a return of cents per day is sarcastically thanked by other miners still in the game, who get very cheap or 'free' electricity.  Other miners are now less likely to quit as someone else will take their share, and it's a race to the bottom from here.

In reality we don't need any miners.  There are 7 million bitcoins in existence which is probably more than half what will ever get generated (early adopters who mined and thought the whole thing to be pointless and deleted their wallet, hard drive crashes, stolen and reformatted computers, deceased bitcoin owners not informing anyone of their wallet, etc).  More than enough bitcoins out there to satisfy speculators.

Except for the fact that the security of the existing bitcoins is directly related to the difficulty level. The mining part isn't there to distribute bitcoins, it's there to keep them safe.

Wow, 203 posts and you still do not know that you need mining for the transactions to go through?

that may be your fuel charge, but depending where you are your fuel charge is about half of your KWH charge... That or you live somewhere with rates that are less than 1/2 those of Austin Tx, and we are know for having some of the lowest rates in texas..  (with the summer surcharge, after the first 500KWH's per month I pay .13XXX in the summer and about .115XX in the winter.. however the first 500KWH's only cost me .095... With my 3 miners I need 10.00/day to break even.. (when you include the increased AC costs, Fan costs, etc.. (the miners alone are about 7.00/day) and thats running my house at 80' F so not really "cold".

I turned off all 3 of my miners on Saturday morning. I'll turn them back on when the price goes back up to $13 or I'll scale down over the next few weeks.. I'm not going t loose 2 or 3 dollors a day just to keep the bitcoin network secure.

Josh

Oh, I know that.  By 'we don't need miners' I mean we don't need hoards generating bitcoins for little to no profit at current prices.  There will always be miners out there mining with very cheap or 'free' electricity, so transactions will always go through.

The hash rate has barely risen over the past few weeks so it seems we're well past the moment where GPU mining is attractive.  Next up is FPGA mining, assuming someone puts together an attractive package.

I'm sure if we ever have another spike where bitcoins surges to 30 dollars we will see a huge difficulty increase where people run to buy every ati card possible.

Yep this is why I created this post. Those who have to pay above 10c per KW for electricity, should start wind down their operation now. The only miners that can remain in the game for the long run, are those that pay zero or almost nothing for electricity. These people will laugh at difficulty rise, price drop, they don't care, they will still turn a profit. You just can't compete with these guys.

See for yourself. Removed for privacy

DROOL!

Holy Snikies thats insane!!!! I need to rent a house in Slidell and load it up with miners.. thats just straight crazy energy pricing there !!!  Looks like your coop doesn't charge transmission rates.... Me = shocked right now :-).

I would call you a lucky bastard, but living Louisiana during this heat-wave probably isn't all that much fun... I'm not a fan of the 105's we keep seeing but at least the humidity is below 40%.. :-)

Josh

DELETED for privacy

For me  equilibrium is about a month away. Do I plan to sell my cards, with that much raw computing power I can think of a few other projects to use them for in small bursts, not 24/7 like they have been. Maybe even play a few games, without worrying about the drop in hash rate.

Here is the guts of a post I was going to make when I noticed this thread and instead of posting a new thread with a very confusing title of "Sky rocket or go Magma diving ?" I decided to add to this thread ..... they are both ultimately about your own personal equilibrium day.




Sorry for the poor beginning, I always seem to start my posts at the middle and work my way towards the beginning and finally get to the end.

Cost of Electricity per Kilowatt hour unit for me $0.25 (fixed)
Power used by my rig per day (about 220W x 3 + misc) circa 750Watts (24*0.750*0.25) = $4.50 (fixed, unless I add more hardware)
Hashed by rig per day on average 1300MH/sec @ current difficulty of 1888786.70535 is 0.692 BTC.

Assuming that 1 BTC is currently worth about $10 this would be $6.92 mined per day. Great news for me, I am printing money, at a profit using the current exchange rate.


If I assume that the value will increase in the future, I could print at a loss today for a possible profit tomorrow.
Or I could assume that BTC is a bad techo/nerd/geek/investment fad and the value will suddenly drop through the floor, and that I am printing money for a loss today and will be hit by an even larger loss tomorrow.

But I am an optimist and am reminded of that old saying "Always buy land, there ain't any more of it coming along any day soon", that is of course unless you live in Hawaii :)
But if you bought land 100 years ago, there is one thing for sure and that is that it will have appreciated in value. And if I bought land today, in 100 years time it will also have appreciate in value. In between there are ups and downs but overall the value it is up. I think of BTC in the same way as minerals, there is a limited fixed supply. There are ways to get more minerals, by bring them back to earth from space, but the extra mass would slowly spiral our planet into the Sun.


So anyhow back to the task at hand lets work backwards. i.e. At what difficulty will the cost of electricity for me  match the BTC being printed per day (at given exchange rates) ?



I have one constant the cost of electricity and two variables: the BTC generated per day on average and the exchange rate of BTC to dollars.
So to balance both sides of an equation I need to reduce the BTC generated per day to match my electricity costs.

$4.50/$10.00 would be 0.45 BTC per day to be operating at no loss and no profit. Or one 50BTC block found on average once every (50BTC/0.45BTC) 111.111111 days (111 days, 2 hours and 40 minutes)

Now here comes the flaw in trying to work out the magic difficulty, this duration of 111.111 days is larger than 14 days, which is the target duration of each new difficulty level (probably closer to 12 or 13 days in reality).
There is also a flaw that the future exchange rate all the way during the total 111.111 days will be constant and at today's assumed exchange rate of $10.00

Even with these two major flaws in the calculation both of which are unpredictable (BTC exchange rate and future difficulty growth[reduction] rate). Is having a flawed number better than having no number ? I would like to see a bad approximation rather than nothing.

The maths is trivial - see https://en.bitcoin.it/wiki/Difficulty#How_soon_might_I_expect_to_generate_a_block.3F I'm just flipping the formula to get a difficulty instead of an average time for a new block.

difficulty =  ( average duration for a block in seconds x hashrate in hashes per second) / 2^32  = ( 9,600,000 x 1,300,000,000 ) / 4,294,967,296 = 2,905,726.433

In my case a difficulty of 2905726.433 at a hash rate of 1.3Ghash/sec would yield one 50BTC block (ignoring block fees) found on average every 111 days, 2 hours and 40 minutes or 0.45BTC per day over 111.111 days (or a 54% increase from the current difficulty). This is as I mentioned assuming a constant exchange rate of $10 and a constant difficulty for 111.111 days of the 2905726.433 difficulty. 111.111 days divided by 14 days is 7.9, so the difficulty should have changed at least eight times during this period. But there is no way to know if the future difficulty will increase or decrease in the short term. In the long term it should increase, with increasingly more efferent future hardware.


Anyhow there is a decision that I will need to make, circa a month from now. Nothing is set in stone, the BTC exchange rate could sky rocket, or go magma diving. And the difficulty rate could increase (or decrease) more rapidly than expected. There should be enough information above to help you workout your own approximation to your own personal equilibrium day.

EDIT: Actually in about a months time it will be getting close to winter where I live, so maybe I will continue to mine for subsided heating. *evil grin* Maybe next spring/summer you will see a drop off in mining, but I think that every winter there will be an increase. Why generate heat without doing something useful to make the heat.

There is an interesting idea, electric BTC radiators. Increase the hash rate to increase the temperature, and throttle the sharerate to lower the temperature. Maybe pulse it 5 minutes generating then allow it 5 minutes of idle, and adjust the on/off ratio to adjust the average room temperature. It would definitely be an idea that would get media attention. Although I don't know if it would be in a good way, or a bad way.

Anyhow here is a simple table, everyone loves tables, well except those who love graphs, bloody hedonists.
The top row is cost in US cents per KWhour of electricity and the first column is the BTC to US dollar exchange rate.

You are either ON the table or OFF the table. This is because "average duration for a block in seconds x hashrate in hashes per second" is a constant for everyone mining. So if you are mining less than the required number of BTC's a day to break even you can not be on the table above. So in my case I am mining 0.69 BTC per day (1.3Ghash/sec) and my electricity cost is 0.45 BTC per day. If I was CPU mining at 10MHash/second generating 0.01 BTC a day with the same power cost of  0.45 BTC per day (or anything greater than 0.01), even though I could read values from the above table, I would be asleep probably in deep space and would definitely not be ON the above table. My current exchange rate ($10 per BTC) and current cost (18.75 cents per hour) of electricity would be there and point to a cell with the difficulty equilibrium answer if I was mining sensibly.

So My rig uses 750 watts per hour or 0.75 KW hours at 25 cents per KW hour unit. Thats 18.75 cents (0.75*25) , the 20 cent column is the closest to my true value and at a current exchange rate of $10 would give the approximate value of 2724118 for 20 cents but if I adjust this (20/18.75*2724118 = 2,905,725.867) I get the same value as above, or as close as makes no difference: 2,905,725.867 instead of 2,905,726.433

There's absolutely no reason to mine at a loss beyond equilibrium, when you can just buy the same amount of BTC for LESS money!!!

That is true, apart from it is not at loss it is heating the house during winter. So in effect it is subsidised heating. But come next spring it will be time, for me anyhow, to stop mining and buy Bitcoins.

I rented a new place.
I pay only 10$ [in local currency] for electricity no matter how much kw's i use  ;D

I am thinking of creating a bitcoin farm  ::)

Start your farm, and see how long that lasts.

Just because they don't care if you run your stove 10 minutes or 20 minutes doesn't mean they don't care if you run a small server farm in the house.

You see, if you exclude things like server farms, electricity use CAN be guessed or estimated. Most people take a shower, make some breakfast, wash 1 weeks' worth of clothes each week, watch a few hours' TV, leave a PC on 24/7, etc.

But if you start a science lab in your basement, and run up a $500/month electric bill trying to re-create Frankenstein's monster, you can bet the landlord WILL notice.

I'm sure that the landlord's first $400 electric bill for your house will cause him to pay you a visit. He's not in the business of losing money.

So..  he will get free electricity for at least a month and possibly more until your landlord goes through proper evition procedure.

So..  he will get free electricity for at least a month and possibly more until your landlord goes through proper eviction procedure?

No sir, here is the thing.
There is no KW-Meters thingy.
The 10$ [50 Egyptian pounds] is a fine for not installing one <= which is totally legal.
And i am fine with that fine :D

If you get "free" electricity. Isnt that in your contract/lease?

No, he have no way of finding how much KW i used.
So i get free electricity and pay a 10$ each month as a fine.

And no that won't change in the -near- future.

Yep, eventually the mining game will be limited to people with free or low flat electricity cost, like newminerr. Basically anyone paying more than 0.5c per KW is going to be looking at mining at a loss in the near future, you just can't compete with people like newminerr that pay $10 flat for unlimited electricity.

If the price stays this low difficulty will drastically lower making it profitable again for quite a few people who do pay electricity.

I hope so.