Topic: What's your shutdown point? (Read 11441 times)

When I'm doing it for profit I can't be bothered when the price is under $4 at the current difficulty, but right now it's getting cold so I'm just switching on as much as I need for heating.

I've got 3 miners, two of them gives 2.7 MH/W and one 2.4 MH/W at peak efficiency measured for the complete system at the wall. If I want more heat it's less, though. Unfortunately I've spent a lot of time testing different combinations of software and settings to get that result and I would be shooting myself in the foot if I just gave it away.

Well when you production cost is $0.00 it is kinda hard to lose money.

OK so then people with cheaper electricity are pretty safe then .

Funny you said that.  My goal for 2012 is FPGA.  No reason to shutdown the GPU rigs though (until they are no longer profitable).

BTW.  No need to sell entire rigs.  I simply swapped out a hodgepodge of GPU (6950, 6970s, 6870s, 6850s) for what I found to be the most efficient setup (3x5970).  It really didn't cost much out of pocket as for some reason the 6950 junk sells at a premium and the 5970s at a discount.  I don't understand it but I was happy to take advantage of it. 

FPGA will be essential for all miners but it likely will take some time.  The price of bitcoin for mining is pretty much meaningless. Bitcoin could go to $30 US tomorrow and as long as miner's felt confident that it was sustainable you would see difficulty rise until the same people who are break even @ $3 USD are break even at $30 USD.  Knowing you are more efficient than the median miner is a comfort.  The price of bitcoin doesn't really matter.

I understand you're argument; you bought most of your hardware recently and are in it for "the long haul".
But the same argument could be applied to FPGAs -- why not go all out and be king of efficiency?
It's basically your strategy on steroids. Spend even MORE money for an even MORE sure-fire ability to beat the competition.

I haven't bought any new hardware in months. That's why I don't own any 5970s. And I'm too old (with a family, kids, job, etc.) to mess with selling all my rigs and starting over with "more efficient" ones. The price of BTC just doesn't justify the time I'd spend.

I'll admit -- I spent some time back in May/June learning about BTC, setting up rigs, researching cards, etc. but back then there was good money to be made. I would never spend that kind of time now just to earn a few bucks a day (assuming 6 GH/s worth of rigs). Fortunately, I learned about BTC & set up my rigs when BTC were worth much, much more.

I'm actually big into efficiency. I just have to balance it with my other requirements (time, money, etc.)
Before Bitcoin, my whole household used about 13 KW/h of electricity a day. I know all the tricks to save electricity.

Believe me, if it were easy, I'd upgrade all my rigs to be more efficient. But I can't spend any time "tinkering" in my Bitcoin room these days; the $3 or $4/day just doesn't justify much time spent.

I bought them mostly used (4 new when Newegg did their sales) for an average of $350 a pop.  Lots of gamers can't afford top of the line.  I would have no problem unloading them for $300 right now.

Of course higher efficiency means I don't need to unload them.

You can't both admit to not caring about efficiency and then complain about how low efficiency means you are close to break even.
Efficiency is king (unless you are using stolen hardware and/or free electricity).

I'm running an average of 3 GPUs per system, all Linux (except one machine which has to be Windows XP).
One thumb drive; the rest are hard drives. All but 2 systems (used for other things) have a single 2 GB RAM stick.
A couple 80+silver PSUs, the rest are 80+ bronze.
All of them have the extra mobo crap turned off; that's easy stuff.
I'm using AMD Athlon II CPUs. Yes, a Sempron uses even less, but it is also (even more) worthless when it comes to resale time. An Athlon II doesn't exactly suck electricity -- especially when it's not running 100%. I'm not CPU mining.

When I built these systems, MH/W efficiency was NOT my only concern. I was also quite concerned with MH/$ at the time. And motherboards with more than 3 PCI-E slots were much more expensive than the $50 and $70 mobos I could get with 3 slots. I was concerned with payoff time. Same for PSUs. Most of my PSUs are 750W, which can handle 3 cards. A 1200W, which can handle 66% more cards, costs about 100% more. Didn't seem worth it to me.

I'm still a bit glad that my most expensive card is a 5870 -- worth around $230 new. Those 5970's are going to be *extremely* hard to sell used, considering they're worth $500. Any gamer with $500 for a new video card is *going* to buy a new card. End of story.

Or you are willing to wait a long time for returns.

Well there isn't anything requiring you to get 1.75 MH/W or worse.  The question of thread (and re-asked by OP) is what is YOUR shutdown point.

Efficiency is king in any commodity based business. Yes generating bitcoins is the perfect commodity based business.

The price/difficulty factor will be set by the median network efficiency.  Given some % of network involves stolen hardware and free electricity that makes efficiency even more important.

A couple questions on your 1.75 MH/W
Running linux?
Using a thumb drive?
Using a Sempron?
Turned off onboard audio, SATA, USB3.0, RAID, etc on MB?
Using high efficiency PSU?
Underclocked CPU & RAM in BIOS?

Still even at 2.0 MH/W (1.75 is just pathetic) and $0.09 electrical cost we are talking $1.30 per BTC.  Anyone who tries can build a rig more efficient than 1.75MH/W.  If you aren't trying then given there are no barriers to entry price is going to move below your production cost.

IMHO if you are not off by now you are either stealing electricity or the most efficient miner alive.

I would like to point out, that most people are getting MUCH, much less than 2.5 MH/W.

I'm getting 1.75 MH/W on average, including the cost of fans (though I haven't calculated it perfectly yet) and all my cards are 58XX series! All the people using 6XXX cards are going to get much less. (And let's face it -- 58XX cards aren't exactly available at Best Buy these days...) Though it's possible most of the 6XXX crowd were among those who quit in the last couple months...I'd be willing to believe that.

2.5 MH/W is what a couple (?) guys are getting, who sold their previous rigs and bought 5970's to create "maximum efficient rigs". But most of us did NOT sell our rigs and buy 5970's when they became available recently.

Well good news is difficulty has fallen by 1/3rd since this thread began.

So current difficult = ~ 1.2M
Hashes per BTC = (difficulty)*(2^32)/(50 BTC per block)
Hashes per BTC = (1,200,000)*(2^32)/50
Hashes per BTC = 103,079,215,104,000
MH/s per BTC = 103,079,215

At 2.5MH / W
kWh per BTC = (MH per BTC) / (MH/s per W * 3600 seconds per hour * 1000 watts per kw)
kWh per BTC = (103,079,215) / ( 2.5 * 60 * 60  * 1000)
kWh per BTC = 11.45

At $0.09 per kWh
USD per BTC = (kwh) * (electrical rate)
USD per BTC =  (11.45) * (0.09)
US per BTC = $1.03

Still $2 BTC would suck unless we see a significant difficulty decline as electricity would be eating up over half of gross revenue.

On edit: To be more universal anyone can calculate their electrical cost of production with this formula.

Electrical Cost per BTC = (electrical rate)*(difficulty in millions)*(2^32)/ ( 50 * (MH/W) * 3600 * 1000)

Simplified (23.86 is magic number which is simplified result of other conversion factors)

( Electrical Cost / BTC ) = (23.86) * (electrical rate)*(difficulty in millions) / (MH/W)

Example:
( Electrical Cost / BTC ) = (23.86) * ($0.09)*(1.2) / (2.5)
( Electrical Cost / BTC ) = $1.03

Getting close to shutting down my miners again. Lower difficulties at $3 was nice.. but sub $2 would hurt.

How you guys fairing?

Difficulty and price was never chicken or the egg. Miners never voted on what btc is worth. They must decide if it's profitable mining by first looking at the price compared to current difficulty and their electricity rate. If is it profitable to mine in you location based on that then you mine. If it's not then you would get a better price from the market offerings. Explain to me how you can do this backwards as in chicken or the egg? If you are thinking hoarding, that is different. That has to do with miners feeling more confident about the future price and not selling immediately putting themselves at risk to possibly gain more.

It's always been one way. And huge people quitting mining or jumping in to mine does not change the btc generation rate. It just changes how many people will have to share the same amount of btc being generated every day. And bitcoin mining was never get rich quick scheme. It's had a short run of couple months from Apr to June rally with large profit margin followed by now 5 month long decline which narrowed it back down to almost nothing. It was just more competitive way to get bitcoin than buying at market price at different junctions of bitcoins history depending on your electricity rate. 

Might I also add that the global hash rate is one way for the miners to vote on how much BTC is worth.
That is huge unintended benefit of mining that many fail to recognize.
If half the miners got up and quit, the price/BTC would defiantly go a lot lower.
The perception that mining follows price is an argument of "what came first, the chicken or the egg?"
As bitcoin continues to mature the relationship between global hashrate and price will become more apparent.

In the future, Bitcoin mining will not be a get rich scheme - it will essentially pay for itself + very little extra.
With patience, it could pay a lot more if the miner holds onto the coins as Bitcoin's value has the potential to rise over time.

To the OP point, my shutdown point was $4.75. Sold two rigs and kept the last one that has dual 6950s. Crossfired them and now have a kick ass Battlefield 3/COD MW3 machine. Will keep it to mine when I'm not gaming. Dual use is the only profit to be gained with coins so low.

That's not how the attack would work. A person in control could reverse transactions THEY send and also prevent block confirmation. So a correct analogy would be:

"I sent you 100 bitcoins then a couple minutes later they disappear from your wallet because I reversed the transaction." Bitcoins aren't being created/destroyed, just recalled.



Since difficulty is tied to hashrate, and the attack vector I'm talking about is the only risk the network runs into when it has low hashrate/difficulty, I think we're still talking about the same thing.

Not in the mood to argue, just disagreeing. Respectfully.

Only?  Would be like saying I have this machine that if you put in a dollar two dollars come out.  However it only doubles money so likely there is no value in it.



That wasn't the statement you corrected though.

Is the network stronger or weaker if difficulty is lower.  I never said it was a huge risk. It isn't much of a risk (at this level of difficulty) because difficulty IS high.   Higher difficulty makes the network stronger.  That was my statement you had a problem with.

Saying it isn't a risk because difficulty is strong is like saying: "We have a lot of cops, so crime is low.  So why have so many cops.  We might as well fire 90% of them because the risk of crime is low." 

I still disagree that this is as big a deal as people are making it out to be. Even if total network hashing power were to drop by 50%, a single attack vector would require 4.2Thash/sec to gain the magical 50% hashing power required to cause trouble on the network. So there's two points to be made there:

- 50%+ control of network only allows the attacker to essentially stop transactions + reverse transactions they send.
- their control of the network would only be through the next difficulty increase which could possibly net them 50,000 bitcoins if they were able to hold the 50%+ control/block generation

It's my opinion that this type of attack would cost too much to coordinate for a single entity, and as a community we can prevent it from happening by a malicious pool owner by making sure we distribute hashing power among pools. There's not much to be gained by this type of attack vs. the cost/effort involved to exploit it.

In an ideal/healthy bitcoin system, difficulty/hashing/price are self-regulated to the point where mining is barely more profitable than just buying coins.

So I think we just have a difference of opinion here on the impact/risk/probability of this type of attack. And that's OK

And that is difficult why?
Because of high difficulty and large number of miners.

If marginal miners quit what happens to aggregate hashing power?

lower difficulty = lower aggregate hashing power = greater threat of 51% attack.

I never said Bitcoin is vulnerable @ 8.5TH/s but those advocating that we need lower difficulty are essentially saying "we need less network security".


Distribution is important but that is a different (but related) issue.

There was nothing wrong with my quote.  Higher difficulty = higher aggregate hashing power.
That is a good thing.  From a network security standpoint how does lower difficulty help Bitcoin?  While it may help you (and me) personally to earn more coins that isn't the point.  The coins exists and incentive to keep the network strong.