Topic: Mining responsibly (Read 4558 times)

Most ATI cards have different voltage settings for 2d, video, and 3d modes.  Many overclocking tools will let you set only the 3d voltage.

It depends on the leakage of your cpu. It won't be the same for all cards even if are the same model.

My 2 HD 6950 are stable at (1.1V, 800MHz stock):
* 0.875V, 600MHz
* 0.930V, 700MHz
* 1.0V, 800MHz
* 1.1V 920MHz

I'm using the 700MHz setting atm, the power drained is very low.

ATI cards will undervolt by default when they run in 2D mode (lower frequency), so it's quite possible. Unfortunately the only way I can do it on the cards I have is by flashing the firmware. If you're lucky Afterburner can do it on the fly.

Right now I am in the same boat you are in.  I am looking to the future and seeing me shutting down in summer (and probably forever as raw mining) as difficulty increases.  Of course an increase in value of Bitcoin could change that.  

Now has anyone experimented with UNDERVOLTING video cards?  I used to have a pentium M ULV that I could undervolt by 25% (no overclocking!) and it ran perfectly.  I increased my battery life by doing this.  

Most cards I assume could take some degree of undervolting just like they take some degree of overclocking.  

Difficulty just increased 52% today, so if your income was 44 BTC/day the previous week it will be about 29 BTC/day now. The new estimate is a 30% increase the next time, and if other people has been as optimistic as you that will only get higher during the week.

Aha, that's the website I was looking for earlier today.  I knew it was in my history somewhere, but searching for "bitcoin" leads to an increasing number of various websites lately >_>

Many thanks, good sir.

all difficulties:
http://nullvoid.org/bitcoin/difficultiez.php

projection of next difficulty increase (only ~2 hours left at this difficulty )
http://nullvoid.org/bitcoin/

Rather than keeping track manually, I'd suggest looking at https://bitcointalksearch.org/topic/difficulty-over-time-data-2345 especially m0mchil's post with the raw data. 

Not based on the difficulty, but related.  Imagine what would happen to the price of gold if it suddenly could be mined for $10 an ounce.

Both the difficulty and the BTC price have a direct relationship to the total number of people interested in Bitcoin.

Bitcoin has a utility value not present in tulips.  People have a fundamental reason to buy BTC unrelated to the chance that they may be worth more later.

(BTW... I dumped a large amount of BTC at MtGox all at once not too long ago just to see what would happen.  Wish I didn't)

I feel safe to assume that if the price of BTC were to fall, it won't stay seriously depressed for much more than the amount of time it takes people to wire new money to MtGox.  The level of interest in BTC is not going away - something seriously bad will have to happen to BTC for the price to fundamentally fall out of the floor.  

Bitcoin is still a "virgin" to a few milestones that are guaranteed to shake it up and flood it with attention and interest:  The day Bitcoin ends up in the news because some website selling drugs, or somebody gets arrested for something where Bitcoin was part of the story.  The American news media would salivate over the headline "website sells marijuana online, cops can't track it, and now the money's unstoppable too".

If BTC runs in the downward direction, I'll be lying in wait to buy a few of them up.  I bet a few others will too.  Unless someone wants to dump millions of them, I feel that even in spite of a dump, they won't be cheap for long.

If people believe the value of BTC's are based on the difficulty of mining them, then you have a positive feedback loop that creates a bubble. The price of tulips cannot increase forever. The advantage of investing in hardware is that you have some residual value left in the hardware if the market goes south, but then you are left with the hassle of selling the hardware.

My advice is upgrade your existing hardware with the best graphics card it can support and enjoy your monster graphics capability and the extra BTC from mining.  If you believe BTC's are worth $100+ each, then save yourself the hassle and by them outright.

Here is the bright flip side:

I recently made postings about how I thought BTC were overvalued.  That's when difficulty was around 25000.  Now, we're looking at more than double that.  That made me think - time to buy up all the BTC I can.

The harder it is to mine, I believe the higher the price of BTC will go.  You can't have 30% and 45% jumps in mining difficulty without a corresponding increase in people becoming interested in Bitcoin and being willing to plop down money for it.  That in turn directly drives demand for BTC and hence price.  I'll put it this way - if everyone wants to throw in the towel on mining and wants to dump their 5970's, I'll offer $400 for them!  Long term, that would be a big steal.

Past threads of this forum echoed the same thing - that the easy days of mining are over - when BTC was ten or twenty cents.  Then BTC jumped to around a dollar, and suddenly mining becomes a literal gold rush.

Here is a real question to ask: when mining is easy, who's gonna go buy BTC at $10 when it can be mined for a nickel?

Take away the easy mining, and the ground floor price has to go up.  No miner in their right mind is going to offer their BTC for a zero return!

The quick return on investments is just about up.  About every 2 weeks the difficulty increases about 20%, meaning, the average BTC mined decreases by about 20%.  Unless the price of BTC jumps sharply again in the next few months, the ROI will decrease to a year or more.  There is also a potential for the price of BTC to run the other direction which could push the ROI further than 2 years out.  Here are the difficulties I recorded.  (Note that I started recording them on 1/26.)

1/26 - 18437.64 ( I don't know how long it was at this difficulty.)
1/27 - 22012.49 ~ 20%
2/8  - 25997.87 ~ 20%
2/18 - 36459.88 (40% increase in difficulty.)

I spent around $2700 for hardware, and I'm guessing the next increase will be sometime around 3/4 or 3/5.  At the current rate of difficulty, I average around 44 BTC per day, so I figure:

2/26 to 3/4 - 44*7 = 308
3/5 to 3/18 - (35.3*14) = 492.8  (2.7+ months to recover 2.7k assuming no change in difficulty and $1 BTC)
3/19 to 4/1 - (28.16*14) = 394.24 (3+ months to recover 2.7k assuming no change in difficulty and $1 BTC)
4/2 to 4/15 - (22.52*14) = 315.39 (4+ months to recover 2.7k assuming no change in difficulty and $1 BTC)
4/16 to 4/29 - (18.02*14) = 252.31 (5+ months to recover 2.7k assuming no change in difficulty and $1 BTC) (my ROI)
4/30 to 5/13 - (14.41*14) = 201.82 (6+ months to recover 2.7k assuming no change in difficulty and $1 BTC)
5/14 to 5/27 - (11.53*14) = 161.42 (8+ months to recover 2.7k assuming no change in difficulty and $1 BTC)
... etc...

So a straight line of 161.42 per month assuming $1 per BTC, it would take 8+ months to pay for my hardware again (not including electric costs).  If you take that one more month out, you are looking at 12+ months to pay for hardware again (not including electric costs).

If I were to do it all over and buy the hardware now, I would not break even until sometime in September (7 months away).  If I were to bet on the price of BTC going up, then it may be sooner.  However, if I think the price will go up again, I would not invest in hardware again, I would purchase BTC's outright.  In fact, had I purchased BTC's instead of $2.7k of hardware, I would have been ahead by about $1k of my current position. (But building those systems was fun!)


Note that I spent around $2700 for the 5 graphic cards and two extra computers.  One 5870 for my existing rig, and 4 5870's on two new systems.  I was not aiming to optimize for khash per $ investment.  I was aiming to create reasonably buildable systems that would not crash.  I also have not over clocked, but I may switch to Ubuntu for added performance.  

Solar pond is an option. 

http://en.wikipedia.org/wiki/Solar_pond

1)  This first line did more damage than good if you are trying to advocate lower power consumption for others.  Now, many people will run out and buy hardware and start mining based upon this projection.

2) Respectfully, the projection may be incorrect.  I have 4 5970's myself so I've done all the math.  Difficulty has shot way up recently with no corresponding increase in bitcoin price. In fact, price has gone down.  The market is volitile enough that if you are holding bitcoins you cannot be sure what they will net you until you sell them.  If bitcoins stay at a dollar and difficulty stays the same you'd be right on.  However, I'd be interested in a calculation where someone accurately projects out the recent exponential increase in difficulty for a month and a half.  But even with you calcs at present difficulty and 5870's being a minimum of $200 (more like $250 w/ tax & shipping but we'll stick w/ $200), how are you building computers for under $290 each considering mobo, adequate power supply, cpu, memory all of that?

For spring and fall closing vents and opening the windows would work but in some areas you would risk overheating your hardware in the summer.

So... Put your mining rigs in the garage or some similarly outside-but-not-really area of your premises...  A guest bedroom where you can close the door and AC vents and open the windows; a garden shed with an extension cord; even on the porch, if you live in an area where you wouldn't expect to come home and find them missing.

Though still admirable to only mine at night, you don't need to increase your air conditioning load.

Do you have a water flatrate included in your rent ?
-> Build a hydro-electric power plant inside your appartment ! ;P

ROI on solar panels is around 20 years for conventional technology.  With the new thin film technology the ROI is approaching 5 years.  These ROI's don't batteries.  Unfortunately, the new thin film technology will not be available for general consumer use for some time.


You've "sparked" a new idea!  If I can save 50% on my electric bill by only using electricity at night, what is to stop me from storing that electricity at night and using it by day!  I once considered solar panels, but the ROI is to large (20 to 25 years).  However, in the process of researching this, I stumbled across Ni-Fe batteries and figured I could store enough energy for may day to day use for about $4k.  (Lead acid batteries cost half as much... so $2k for conventional car batteries.)  So, if my average electric bill is $200 per month, I could save close to $1k per year in electric costs.  This gives a ROI of 4 to 5 years if I were to take the nighttime rate electricity and use it for day time use.  Also, the Ni-Fe batteries last for 50 years and even then can be rejuvenated with cheap readily available chemicals.  Basically, they last forever, as opposed to 5 years for lead acid batteries.

(http://www.beutilityfree.com/Electric/Ni-Fe)

Agreed, solar.