Two great articles
http://www.coindesk.com/microscope-economic-environmental-costs-bitcoin-mining/
http://bitcoinmagazine.com/14282/mining-2/
Topic: Break even difficulty by hardware efficiency (power cost = value of BTC) (Read 18881 times)
June 21, 2014, 08:44:49 AM
May 25, 2014, 04:14:21 AM
great thread 
May 18, 2014, 11:22:26 PM
interesting or bleak?
With the quickly increasing difficulty and the increasing number of ASIC creators and more efficient ASICs, I'll go with the latter.
May 18, 2014, 10:29:39 PM
Getting closer. The break even points in the OP are based on $100 exchange rate. At $400 USD per BTC it would be around 100 billion difficulty, at $500 it is 125 billion. Things will get real interesting over the next six months or so.
May 18, 2014, 09:49:37 PM
So it looks pretty safe to say that the network will be less than 250 PH/s using the 28nm technology at the current prices.
The replacement cost is calculated based on all hashing power coming from a single design so it shouldn't be taken as likely but more an upper bound. I think there might be an error in the math as well.
The cost to reach the "break even point" is based on current cost so I don't find it too useful, because cost will decline significantly. The network will approach a small margin below the break even point. It is only a matter of time. You can consider that the equilibrium point, we have seen it occur in the GPU era for a couple of years. If more efficient hardware is produced then the hashrate will rise, if the exchange rate rises then the hashrate will rise, as mining moves to lower cost areas the hashrate will rise, if the exchange rate falls the hashrate will fall. However the network will approach and stay close to a margin below the break even point. How close? It really depends on miners and what risk they are taking.
Are we now entering the "equilibrium point" for the 28nm ASIC era?
March 17, 2014, 10:09:09 AM
I have created the following thread using the numbers calculated in this thread:
https://bitcointalksearch.org/topic/projected-minimum-cost-per-btc-over-the-next-year-518111
Question: why is this specification https://bitcointalksearch.org/topic/annasicminer-publicly-looking-for-potential-customerspartners-for-new-chips-438359 (0.2 - 0.35 J/GH) not being taken into account in the OP? Or is it? Perhaps they are quoting power at the chip and you are quoting power (0.8 J/GH) at the wall?
https://bitcointalksearch.org/topic/projected-minimum-cost-per-btc-over-the-next-year-518111
Question: why is this specification https://bitcointalksearch.org/topic/annasicminer-publicly-looking-for-potential-customerspartners-for-new-chips-438359 (0.2 - 0.35 J/GH) not being taken into account in the OP? Or is it? Perhaps they are quoting power at the chip and you are quoting power (0.8 J/GH) at the wall?
They almost certainly are quoting the chip. We only care about wattage at the wall. Also not sure if that chip has been released yet, it certainly wasn't at the time of the OP. The OP refers to ASICMiner 130nm the linked post was for 40nm. Most companies miss their target numbers. I would look at them with a good dose of skepticism until you see reports of users in the field getting x W/GH AT THE WALL (i.e. killawatt meter of equivalent).
March 17, 2014, 01:02:51 AM
I have created the following thread using the numbers calculated in this thread:
https://bitcointalksearch.org/topic/projected-minimum-cost-per-btc-over-the-next-year-518111
Question: why is this specification https://bitcointalksearch.org/topic/annasicminer-publicly-looking-for-potential-customerspartners-for-new-chips-438359 (0.2 - 0.35 J/GH) not being taken into account in the OP? Or is it? Perhaps they are quoting power at the chip and you are quoting power (0.8 J/GH) at the wall?
https://bitcointalksearch.org/topic/projected-minimum-cost-per-btc-over-the-next-year-518111
Question: why is this specification https://bitcointalksearch.org/topic/annasicminer-publicly-looking-for-potential-customerspartners-for-new-chips-438359 (0.2 - 0.35 J/GH) not being taken into account in the OP? Or is it? Perhaps they are quoting power at the chip and you are quoting power (0.8 J/GH) at the wall?
without fail, every single company that tried to forecast their power consumption has got it wrong (usually at least 20-30% underestimating), so its always best to try it after the chips come back from the fab to see what the actual power consumption is.
also, as you say, most power we measure is at the wall... because there are significant losses due to dc/dc power conversion, and ac/dc power conversion, and cooling systems, and host computers, etc... so its best to compare like for like and compare them all at the wall.
March 16, 2014, 08:26:26 AM
I have created the following thread using the numbers calculated in this thread:
https://bitcointalksearch.org/topic/projected-minimum-cost-per-btc-over-the-next-year-518111
Question: why is this specification https://bitcointalksearch.org/topic/annasicminer-publicly-looking-for-potential-customerspartners-for-new-chips-438359 (0.2 - 0.35 J/GH) not being taken into account in the OP? Or is it? Perhaps they are quoting power at the chip and you are quoting power (0.8 J/GH) at the wall?
https://bitcointalksearch.org/topic/projected-minimum-cost-per-btc-over-the-next-year-518111
Question: why is this specification https://bitcointalksearch.org/topic/annasicminer-publicly-looking-for-potential-customerspartners-for-new-chips-438359 (0.2 - 0.35 J/GH) not being taken into account in the OP? Or is it? Perhaps they are quoting power at the chip and you are quoting power (0.8 J/GH) at the wall?
March 15, 2014, 08:18:38 AM
I can see the difficulty becoming way to high regardless if you have 1 TH, theres so many companies now looking to sell bitcoin miners
March 14, 2014, 11:16:45 PM
Maybe I am just too tired right now to recognize the obvious... but these calculations are assuming a mining reward of 25 BTC per block right? (and ignoring transaction fees I guess)
February 07, 2014, 08:47:18 AM
Hey, kWh is kW times hours NOT kW per hour (like fingernails on a blackboard to an EE)
Also: 0.415 kW not 0.415 kWh - you multiplied by 24 h later to get kWh
Also: 0.415 kW not 0.415 kWh - you multiplied by 24 h later to get kWh
Splitting hairs, it's still valid to work it out as 0.415 kW per hour before you multiply it by 24hrs. (kWh = kW x 1h)
By the way the reason for the bad BTC price seems to be a problem with MtGox.
https://www.mtgox.com/press_release_20140207.html
I wonder how they'll explan it on Monday? Hacked? Liquidity? Manipulation? Other?
February 07, 2014, 08:31:20 AM
Hey, kWh is kW times hours NOT kW per hour (like fingernails on a blackboard to an EE)
Also: 0.415 kW not 0.415 kWh - you multiplied by 24 h later to get kWh
Also: 0.415 kW not 0.415 kWh - you multiplied by 24 h later to get kWh
February 07, 2014, 06:14:55 AM
Work in kWh (1000 watts per hour), that's what you pay your power company in, not J.
My 6x Bitfury Hex16B mine 270GH using 415W @ 11p per kWh
The math is simple...
0.415kWh x 11p x 24hrs = 109.56p (£1.10) power cost per day.
The exchange rate is pretty bad today* 436.6 GBP (713.1 USD) per BTC and they mine about 0.06BTC per day...
0.06 x £436.60 = £26.20 mining revenue per day.
So now you take one away from the other
£26.20 - £1.10 = £25.10 mining profit per day (excluding equipment costs)
The miners cost 6x £415 = £2490 / £25.10 mining profit per day = 99 days to break even if you bought it today.
Keep all your receipts in your local currency not BTC, in case you ever need to explain to the tax man where your income is coming from!
*Seems to be a problem at MtGox http://www.marketwatch.com/story/bitcoin-drops-sharply-as-mt-gox-halts-withdrawals-2014-02-07?dist=beforebell
My 6x Bitfury Hex16B mine 270GH using 415W @ 11p per kWh
The math is simple...
0.415kWh x 11p x 24hrs = 109.56p (£1.10) power cost per day.
The exchange rate is pretty bad today* 436.6 GBP (713.1 USD) per BTC and they mine about 0.06BTC per day...
0.06 x £436.60 = £26.20 mining revenue per day.
So now you take one away from the other
£26.20 - £1.10 = £25.10 mining profit per day (excluding equipment costs)
The miners cost 6x £415 = £2490 / £25.10 mining profit per day = 99 days to break even if you bought it today.
Keep all your receipts in your local currency not BTC, in case you ever need to explain to the tax man where your income is coming from!
*Seems to be a problem at MtGox http://www.marketwatch.com/story/bitcoin-drops-sharply-as-mt-gox-halts-withdrawals-2014-02-07?dist=beforebell
February 07, 2014, 02:25:16 AM
With new 10MW DataCenter KNC can add another 10-15 PHash to network.
November 25, 2013, 06:03:39 PM
Yeah saw that but if the 20nm is a FPGA HardCopy ASIC it will be no faster or efficient than a genuine 28nm ASIC.
I wonder when we will see Bitfury Strikes Again?!
I wonder when we will see Bitfury Strikes Again?!
November 25, 2013, 10:18:55 AM
November 25, 2013, 09:45:13 AM
I have to wonder..
KNC 28nm 1.1 [3] 18,250 130.6 <-- if this is a 28nm FPGA HardCopy not a true 28nm ASIC?
Bitfury 55nm 0.9 23,300 166.8
http://www.altera.co.uk/devices/asic/hardcopy-asics/hardcopy-v/hcv-index.jsp
KNC 28nm 1.1 [3] 18,250 130.6 <-- if this is a 28nm FPGA HardCopy not a true 28nm ASIC?
Bitfury 55nm 0.9 23,300 166.8
http://www.altera.co.uk/devices/asic/hardcopy-asics/hardcopy-v/hcv-index.jsp
November 24, 2013, 11:12:28 PM
So, now $1000 /BTC, the most efficient miner exists today, can profitably hash till the network reaches, 1668 PHash. Awesome. 
I don't think we're going cross this limit anytime in 2014.
I don't think we're going cross this limit anytime in 2014.
I'm excited.
November 24, 2013, 10:51:50 PM
So, now $1000 /BTC, the most efficient miner exists today, can profitably hash till the network reaches, 1668 PHash. Awesome.
I don't think we're going cross this limit anytime in 2014.
I don't think we're going cross this limit anytime in 2014.
October 25, 2013, 09:43:08 AM
1 HP ≈ 746 watts
1 HP ≈ Burning 641 calories an hour
300 calories = 75kg person, cycling at <10 mph, for an hour = (746/641) * 300 watts = 349.14 watts = 349.14 Joules/second = 384 GHash ≈ 2 bitcents @ 3 PHash ≈ $4 @ $200 per BTC
So, with an appropriate exercise machine attached to a miner, your 1 hour workout would also give you some Bitcoin rewards.
1 HP ≈ Burning 641 calories an hour
300 calories = 75kg person, cycling at <10 mph, for an hour = (746/641) * 300 watts = 349.14 watts = 349.14 Joules/second = 384 GHash ≈ 2 bitcents @ 3 PHash ≈ $4 @ $200 per BTC
So, with an appropriate exercise machine attached to a miner, your 1 hour workout would also give you some Bitcoin rewards.
I can see armies of prisoners in china being forced to pedal bikes attached to PCs with massive piles of discarded block erupters right now.
http://www.forbes.com/sites/insertcoin/2011/06/02/chinese-prisoners-forced-to-farm-world-of-warcraft-gold/
October 25, 2013, 02:59:04 AM
1 HP ≈ 746 watts
1 HP ≈ Burning 641 calories an hour
300 calories = 75kg person, cycling at <10 mph, for an hour = (746/641) * 300 watts = 349.14 watts = 349.14 Joules/second = 384 GHash ≈ 2 bitcents @ 3 PHash ≈ $4 @ $200 per BTC
So, with an appropriate exercise machine attached to a miner, your 1 hour workout would also give you some Bitcoin rewards.
1 HP ≈ Burning 641 calories an hour
300 calories = 75kg person, cycling at <10 mph, for an hour = (746/641) * 300 watts = 349.14 watts = 349.14 Joules/second = 384 GHash ≈ 2 bitcents @ 3 PHash ≈ $4 @ $200 per BTC
So, with an appropriate exercise machine attached to a miner, your 1 hour workout would also give you some Bitcoin rewards.
Then you would have to order pizza for BTC to replenish calories lost on mining. Just like an ordinary coal miner :-)
October 25, 2013, 02:55:00 AM
1 HP ≈ 746 watts
1 HP ≈ Burning 641 calories an hour
300 calories = 75kg person, cycling at <10 mph, for an hour = (746/641) * 300 watts = 349.14 watts = 349.14 Joules/second = 384 GHash ≈ 2 bitcents @ 3 PHash ≈ $4 @ $200 per BTC
So, with an appropriate exercise machine attached to a miner, your 1 hour workout would also give you some Bitcoin rewards.
1 HP ≈ Burning 641 calories an hour
300 calories = 75kg person, cycling at <10 mph, for an hour = (746/641) * 300 watts = 349.14 watts = 349.14 Joules/second = 384 GHash ≈ 2 bitcents @ 3 PHash ≈ $4 @ $200 per BTC
So, with an appropriate exercise machine attached to a miner, your 1 hour workout would also give you some Bitcoin rewards.
October 25, 2013, 12:46:52 AM
According to thegenesisblock.com none of the current machines will pay for themselves.... What do you say about that?
May be fiat-in can be paid out as fiat-out. $5k miner + $1k electricity in >> $6k out. I'm sure no one is getting their BTC out; means 100 btc in >> 50 btc. You should consider yourself lucky if exchange rate was doubled.
Now, hoarding becomes a better investment than investing in miners..
October 25, 2013, 12:35:13 AM
According to thegenesisblock.com none of the current machines will pay for themselves.... What do you say about that?
October 24, 2013, 11:38:15 PM
By taking $$$ out of equation, your calculations are based on, 1 mBTC/kWh ($0.1 per kWh, at $100 per BTC), or 500 uBTC/kWh ($0.1 per kWh, at $200).
If some solar / wind (or horse-powered) power companies jump into Bitcoin mining and offer electricity priced in BTC, then Bitcoin mining would look entirely different.
These companies can list their kW shares in an exchange for trading, like cex.io GHS.
I am so looking forward to see this happen.. my prediction / guesstimate; this is going to happen in 2014-Q4.
May be 22nm / 20nm ASICs could slow this down for another quarter, or two.
If some solar / wind (or horse-powered) power companies jump into Bitcoin mining and offer electricity priced in BTC, then Bitcoin mining would look entirely different.
These companies can list their kW shares in an exchange for trading, like cex.io GHS.
I am so looking forward to see this happen.. my prediction / guesstimate; this is going to happen in 2014-Q4.
May be 22nm / 20nm ASICs could slow this down for another quarter, or two.
October 24, 2013, 10:18:03 PM
At 0.9 Joules / GHash, 10 PHash server farm requires, 2.5 kWh?!
I must be wrong somewhere.
Is it 0.9 joules / second / GHash?
In that case, it would be 9000 kWh.
I must be wrong somewhere.
Is it 0.9 joules / second / GHash?
In that case, it would be 9000 kWh.
kWh is a measure of energy (power over time).
kW is a measure of power.
1 kW for 1 hour = 1 kWh
10 PH/s farm
0.9 J/GH * 10,000,000 GH/s = 9,000,000 J/s = 9,000,000 W = 9,000 kW
So yes 9,000 but it is kW not kWh. Now if you ran your 9,000 kW farm for 1 hour it would use 9,000 kWh of energy.
In a year that 9,000 kW farm would use 9,000 * 24 * 365 = 788 million kWh. At $0.10 per kWh that would be $78.8 million in energy cost.
If the Joule conversion confuses you 0.9 J/GH = 0.9 W/GH/s but that looks ugly.
October 24, 2013, 10:07:42 PM
At 0.9 Joules / GHash, 10 PHash server farm requires, 2.5 kWh?!
I must be wrong somewhere.
Is it 0.9 joules / second / GHash?
In that case, it would be 9000 kWh.
I must be wrong somewhere.
Is it 0.9 joules / second / GHash?
In that case, it would be 9000 kWh.
October 24, 2013, 01:29:40 PM
That would put the reported DC output of the module higher (430W ) than the computing DC input (385W). Something isn't correct. So either your wall wattage numbers or the output reported by the VRM is incorrect, they both can't be right. Under ideal conditions (no cooling or host power consumption), 90% DC efficiency, 93% ATX PSU efficiency. 430W output would mean (430/(0.9*0.93) = 513W) >500W at the wall.
next time I'll go to the colo I will measure at the wall wattage again.
Quote
Power Supply model.
I don't remember if I've already said mine is a cooler master v850.
Has anyone run firmware 0.97 with the bertmod files? If so, I'll run this and give you guys watts at the wall vs watts output at the VRM.
http://forum.kncminer.com/forum/main-category/main-forum/6183-bertmod-0-2-unofficial-firmware-mod-feedback-thread?p=9442#post9442
It works.
October 24, 2013, 12:54:50 PM
That would put the reported DC output of the module higher (430W ) than the computing DC input (385W). Something isn't correct. So either your wall wattage numbers or the output reported by the VRM is incorrect, they both can't be right. Under ideal conditions (no cooling or host power consumption), 90% DC efficiency, 93% ATX PSU efficiency. 430W output would mean (430/(0.9*0.93) = 513W) >500W at the wall.
next time I'll go to the colo I will measure at the wall wattage again.
Quote
Power Supply model.
I don't remember if I've already said mine is a cooler master v850.
Has anyone run firmware 0.97 with the bertmod files? If so, I'll run this and give you guys watts at the wall vs watts output at the VRM.
October 22, 2013, 02:53:10 AM
That would put the reported DC output of the module higher (430W ) than the computing DC input (385W). Something isn't correct. So either your wall wattage numbers or the output reported by the VRM is incorrect, they both can't be right. Under ideal conditions (no cooling or host power consumption), 90% DC efficiency, 93% ATX PSU efficiency. 430W output would mean (430/(0.9*0.93) = 513W) >500W at the wall.
next time I'll go to the colo I will measure at the wall wattage again.
Quote
Power Supply model.
I don't remember if I've already said mine is a cooler master v850.
October 22, 2013, 01:58:42 AM
Can we update to $200/BTC now?
October 21, 2013, 06:26:24 PM
Thanks for the datapoints. It is strange the reported DC/DC output doesn't change.
~430W out regardless however the input wattage changes significantly.
Need to make some assumptions but lets say the 6 fans use 6W ea (someone can look at the fan sticker) and the host uses another 5W. So balance of system is ~40W @ 12VDC. Lets also assume your PSU is 90% efficient at
v0.90 = 1650W @ 220VAC ~= 1485W DC @ 12VDC (1485 - 40)/2 = 722W.
v0.95 = 946W @ 220VAC ~= 850W DC @ 12VDC (850-40)/2 = 405W.
v0.90 VRM In: 722W Out: 433W Efficiency: 60% OUCH
v0.95 VRM In: 405W Out: 438W Efficiency: IMPOSSIBLE.
So either your numbers or incorect or the output reported by the VRM is incorrect. It is possible the PSU efficiency in the second case was slighly higher (say 92%) and the host power usage is less but those numbers don't change things significantly. Looking at it the other way the VRM is at most 90% efficient. 438W out = 486W in (@ 90 efficiency) or 962W @ 12VDC for both. Even with no fans or host wattage the AC wattage doesn't match the reported VRM output wattage even under ideal conditions (90% efficiency @ VRM, 92% efficiency @ PSU).
~430W out regardless however the input wattage changes significantly.
Need to make some assumptions but lets say the 6 fans use 6W ea (someone can look at the fan sticker) and the host uses another 5W. So balance of system is ~40W @ 12VDC. Lets also assume your PSU is 90% efficient at
v0.90 = 1650W @ 220VAC ~= 1485W DC @ 12VDC (1485 - 40)/2 = 722W.
v0.95 = 946W @ 220VAC ~= 850W DC @ 12VDC (850-40)/2 = 405W.
v0.90 VRM In: 722W Out: 433W Efficiency: 60% OUCH
v0.95 VRM In: 405W Out: 438W Efficiency: IMPOSSIBLE.
So either your numbers or incorect or the output reported by the VRM is incorrect. It is possible the PSU efficiency in the second case was slighly higher (say 92%) and the host power usage is less but those numbers don't change things significantly. Looking at it the other way the VRM is at most 90% efficient. 438W out = 486W in (@ 90 efficiency) or 962W @ 12VDC for both. Even with no fans or host wattage the AC wattage doesn't match the reported VRM output wattage even under ideal conditions (90% efficiency @ VRM, 92% efficiency @ PSU).
sorry but I think I didn't explain the situation properly (I'm not a native english speaker, plus my eng is a little bit rusty).
I own 2 jupiters and all the ampere measurments reported take into account both of them.
With 0.90 together the 2 jups drain 7.5 ampere @ 220V. For each of them at the wall I get more or less 825 Watts .
Unfortunately I didn't get a chance to use bertmod while using 0.90 (It didn't even exist maybe).
With 0.95 together the 2 jups drain 4.4 Ampere @ 220 V. circa 473 Watts each.
The ouput I get from bertmod reported in the prev post is related to the jupiters while running fw 0.95 (jup 1 = 438 , jup 2 = 433)
edit: fix a few typos
October 21, 2013, 06:00:35 PM
On edit: fixed misunderstanding of numbers reported. Both output DC numbers are v0.95. There are no DC output numbers for v0.90.
Need to make some assumptions but lets say the 6 fans use 6W ea (someone can look at the fan sticker and let me know wattage or amps) and the host uses another 5W. That puts the balance of the system at ~40W. Lets also assume your PSU is 90% efficient at 220V pretty reasonable for 80 Plus Gold unit over most of its operating range (208V-240V tends to be 1% to 2% more efficient than 120V).
v0.95 = 473W @ 220VAC ~= 425W DC @ 12VDC (425-40) = 385W Input for VRM (96W per module).
That would put the reported DC output of the module higher (430W ) than the computing DC input (385W). Something isn't correct. So either your wall wattage numbers or the output reported by the VRM is incorrect, they both can't be right. Under ideal conditions (no cooling or host power consumption), 90% DC efficiency, 93% ATX PSU efficiency. 430W output would mean (430/(0.9*0.93) = 513W) >500W at the wall.
Still lets look at the wall efficiency.
v0.90 825W / 495 GH/s = 1.7 J/GH OUCH.
v0.95 473W / 495 GH/s = 1.0 J/GH
If anyone else has KNC datapoints please provide the following:
KNC Mode.
Average hashrate.
Wattage at the wall.
Firmware version.
# of VRMs present (4 or 8 ).
Power Supply model.
Mains Voltage (120V, 208V, 220, 240V, etc).
Thanks.
October 21, 2013, 04:34:17 PM
Update (improved) KNC to 1.0 J/GH based on reported 455W @ 12 VDC running at 502 GH/s. Assume 90% AC power supply efficiency results in estimated 505W at the wall. Also moved KNC into the "actual devices" category. Bitfury still holds the efficiency crown though.
https://bitcointalksearch.org/topic/m.3307091
On edit: revised slightly to 1.1 J/GH. The 455W was the output (0.75V) of the DC to DC converter. Based on GE spec sheet the input current would be ~10% higher. That puts it closer to 560W at the wall. This appears to be inline with other customer reports using kill-a-watt type meters. On some firmware and with the 4 VRM model the wattage is significantly higher ~1.3 J/GH.
https://bitcointalksearch.org/topic/m.3307091
On edit: revised slightly to 1.1 J/GH. The 455W was the output (0.75V) of the DC to DC converter. Based on GE spec sheet the input current would be ~10% higher. That puts it closer to 560W at the wall. This appears to be inline with other customer reports using kill-a-watt type meters. On some firmware and with the 4 VRM model the wattage is significantly higher ~1.3 J/GH.
just to add a few datapoints.
I own 2 jups both with a 8 VRMs per PCB and a faulty asic module with a die with all 48 cores disabled each.
At the pool I got something like 490-95 GH/s each.
with fw 0.90 the power consumption combined was 7.5 Ampere @ 220 V at the wall --> 1650 Ws = 825 Ws each.
with fw 0.95 the hashrate remains the same and I got 4.4 Ampere @ 220 V at the wall --> 946 Ws = 473 Ws each.
bertmod reports that ASIC slots total DC/DC power output is 433 and 438 Ws respectively.
I user the same PSU model for both: Cooler Master V850.
October 21, 2013, 01:14:27 PM
I think a likely outcome of corporate hashrate monopolisation would be the death of crypto. It's just not a viable way to go because it would remove fundamental human interest in the entire concept in these early years. The technologists of the world, who are you and I, would leap off in droves, and the concept would die. Because we make it, in the end.
I don't doubt that big capital is putting more in as time goes on, but if they go too far and take over what isn't settled yet, they'll end up with nothing.
The thing about Bitcoin is that so far (and probably for quite a long time yet to come) it is an alternative, we all can just dump it at any time and resume our normal fiat lives, should we become sufficiently disillusioned with it for any reason.
I don't doubt that big capital is putting more in as time goes on, but if they go too far and take over what isn't settled yet, they'll end up with nothing.
The thing about Bitcoin is that so far (and probably for quite a long time yet to come) it is an alternative, we all can just dump it at any time and resume our normal fiat lives, should we become sufficiently disillusioned with it for any reason.
October 13, 2013, 07:08:37 AM
October 09, 2013, 10:20:48 AM
Update (improved) KNC to 1.0 J/GH based on reported 455W @ 12 VDC running at 502 GH/s. Assume 90% AC power supply efficiency results in estimated 505W at the wall. Also moved KNC into the "actual devices" category. Bitfury still holds the efficiency crown though.
https://bitcointalksearch.org/topic/m.3307091
On edit: revised slightly to 1.1 J/GH. The 455W was the output (0.75V) of the DC to DC converter. Based on GE spec sheet the input current would be ~10% higher. That puts it closer to 560W at the wall. This appears to be inline with other customer reports using kill-a-watt type meters. On some firmware and with the 4 VRM model the wattage is significantly higher ~1.3 J/GH.
https://bitcointalksearch.org/topic/m.3307091
On edit: revised slightly to 1.1 J/GH. The 455W was the output (0.75V) of the DC to DC converter. Based on GE spec sheet the input current would be ~10% higher. That puts it closer to 560W at the wall. This appears to be inline with other customer reports using kill-a-watt type meters. On some firmware and with the 4 VRM model the wattage is significantly higher ~1.3 J/GH.
September 24, 2013, 05:53:06 PM
September 22, 2013, 10:24:32 AM
Quote
Why did you make this?
I find it silly people have projections with difficulty going to 1 trillion or more (I think the highest I have seen is 200 trillion). At a mere 300 billion difficulty all current and proposed mining devices would be operated at a significant loss (assuming $100 exchange rate & $0.10 electrical rate). So at 1 trillion in difficulty even a Cointera rig would be converting $4 in electricity into $1 in Bitcoins. Anyone think that is likely? Significantly higher difficulty is going to require either more efficient hardware, a higher exchange rate, or the average cost of power for the network to decline.
I find it silly people have projections with difficulty going to 1 trillion or more (I think the highest I have seen is 200 trillion). At a mere 300 billion difficulty all current and proposed mining devices would be operated at a significant loss (assuming $100 exchange rate & $0.10 electrical rate). So at 1 trillion in difficulty even a Cointera rig would be converting $4 in electricity into $1 in Bitcoins. Anyone think that is likely? Significantly higher difficulty is going to require either more efficient hardware, a higher exchange rate, or the average cost of power for the network to decline.
Interesting numbers. We can understand that home / basement / garage based miner's time has come to an end, or will end soon. Considering a data center environment, where hosting / rent is calculated based on 'Rack unit', I am wondering how these numbers would be calculated.
Total network hashrate is still a unpredictable. Is it possible to say an investment amount (miner + hosting cost) will break even in 6 months?
Not necessarily.
Home based miners don't calculate costs for manpower/administration, will often not use AC (which doubles the electricity price) and might more easily mine on a very thin or negative margin. And then there's an army of people with no (self paid) electricity costs at all.
But yes, those groups probably won't ever reach a percentage of the total mining power as in CPU or GPU times..
Ente
September 21, 2013, 05:05:14 AM
Quote
Why did you make this?
I find it silly people have projections with difficulty going to 1 trillion or more (I think the highest I have seen is 200 trillion). At a mere 300 billion difficulty all current and proposed mining devices would be operated at a significant loss (assuming $100 exchange rate & $0.10 electrical rate). So at 1 trillion in difficulty even a Cointera rig would be converting $4 in electricity into $1 in Bitcoins. Anyone think that is likely? Significantly higher difficulty is going to require either more efficient hardware, a higher exchange rate, or the average cost of power for the network to decline.
I find it silly people have projections with difficulty going to 1 trillion or more (I think the highest I have seen is 200 trillion). At a mere 300 billion difficulty all current and proposed mining devices would be operated at a significant loss (assuming $100 exchange rate & $0.10 electrical rate). So at 1 trillion in difficulty even a Cointera rig would be converting $4 in electricity into $1 in Bitcoins. Anyone think that is likely? Significantly higher difficulty is going to require either more efficient hardware, a higher exchange rate, or the average cost of power for the network to decline.
Interesting numbers. We can understand that home / basement / garage based miner's time has come to an end, or will end soon. Considering a data center environment, where hosting / rent is calculated based on 'Rack unit', I am wondering how these numbers would be calculated.
Total network hashrate is still a unpredictable. Is it possible to say an investment amount (miner + hosting cost) will break even in 6 months?
September 16, 2013, 01:10:13 PM
Good list.
Bitfury is somewhat better than that though, I measured +116W for adding a 150GH/s Bitfury unit to my rig. Gold-rated 1300W PSU at 440W@230V total now so I'd assume about 90% efficiency. Anyway, <0.8J/GH at the wall for Bitfury. Another guy had similar results.
Bitfury is somewhat better than that though, I measured +116W for adding a 150GH/s Bitfury unit to my rig. Gold-rated 1300W PSU at 440W@230V total now so I'd assume about 90% efficiency. Anyway, <0.8J/GH at the wall for Bitfury. Another guy had similar results.
Was this on the reduced speed boards? The reason I ask is if the boards intended for 25 GH/s are getting 21 GH/s then they are essentially underclocked (even if unwillingly) and that will improve efficiency.
Was the 116W increase at the wall?
Thanks for the thread, D&T.
BTW, as I'm sure you know, if voltage of the chips is constant, efficiency is roughly independent of frequency(hashrate). While there is a very slight increase of efficiency at lower hashrate, due to better efficiency of DC/DC and main PS, we are talking of single digit percentages only.
Indeed if the voltage can be adjusted, and if you do this properly, efficiency is roughly inverse dependent with the square of frequency(hashrate). For a 10% increase in hashrate you pay a ~33% penalty in power consumption.
On 4 modules Avalon, where voltage is fixed, I have ~8.4J/GH@256MHz degrading continuously to ~8.6J/GH@360MHz. I assume efficiency is a little higher at lower frequencies due to better efficiency at lower power of the DC-DC converters and main PS.
On the other hand, on a 2x7970 rig, at the wall, including motherboard and HDD power consumption:
465 J/GH (2.15 MH/J) @ nominal voltages (1.175 and 1.112V) = 1300MH/s, 605W
322 J/GH (3.10 MH/J) @ undervolt 0.95V = 1100Mh/s, 355W
September 14, 2013, 05:30:47 PM
CoinCraft @ 0.6J/GH - not seen the orignial announcement (if there was one) but zefir announced that he was distributing them here: https://bitcointalksearch.org/topic/closed-bitmine-coincraft-a1-28nm-chip-distribution-diy-support-294235
roy
EDIT: Oh, this is BitMine. And that's 0.6J/GH at chip level....
Based on this page a CoinDesk system is 4.7J/GH is turbo mode and 0.53J/GH in low-power mode. Unfortunately they don't quote power at the wall in 'normal mode' (This is the mode which is 0.6J/GH at the chip, I think - and the mode on which the stated hash speeds are calculated)
roy
EDIT: Oh, this is BitMine. And that's 0.6J/GH at chip level....
Based on this page a CoinDesk system is 4.7J/GH is turbo mode and 0.53J/GH in low-power mode. Unfortunately they don't quote power at the wall in 'normal mode' (This is the mode which is 0.6J/GH at the chip, I think - and the mode on which the stated hash speeds are calculated)
September 12, 2013, 05:31:33 PM
Rising difficulty doesn't give you a hard stop in BTC production though - it just makes your mining more expensive in terms of electricity per BTC - and if your electricity is free, you can keep going until either you, or your miners die, can't you?
Yes, but if you can sell your power hungry 110 nm ASICs and one month later buy far more efficient 28 nm ASICs, why not get a far higher return? It's all about finding another fool to hold the hot (cold?) potato
September 12, 2013, 05:19:24 PM
Even with free power, some people ponied up more BTC for an ASIC (Batch #3 from Avalon for example) than it can ever generate due to the rise in difficulty. I guess some will be so stubborn to let the machines run to recover whatever they can.
Rising difficulty doesn't give you a hard stop in BTC production though - it just makes your mining more expensive in terms of electricity per BTC - and if your electricity is free, you can keep going until either you, or your miners die, can't you?
September 12, 2013, 03:39:15 PM
One other reason why people might rationally mine with hardware that is apparently no longer cost-effective: if they want the heat output anyway.
I've been thinking about this a bit more....
Perhaps the very long term equilibrium is that the electricity costs of mining will be somewhat, or maybe even substantially higher than the returns...
The bulk of mining will be done, for example, in combination with electrical heating (with the coins defraying a part of the electricity cost) and other similar circumstances. The idea of mining and throwing away the heat output (or even worse, using more energy for AC) will be reserved only for the few who have the absolute cheapest electricity, if it happens at all.
That end game is probably a decade or two off, though
roy
September 11, 2013, 02:00:41 PM
One other reason why people might rationally mine with hardware that is apparently no longer cost-effective: if they want the heat output anyway.
For people who use electric heating, why not run an old miner instead of a conventional electric heater? One could even imagine die hard enthusiasts hooking up miners to thermostats and time switches to supply heat when needed.
What would be a decent 1kW miner to buy, as a space heater, I wonder? This time next year I imagine I will be able to pick up a 4-module Avalon for little more than the cost of a fan heater :-)
-roy
For people who use electric heating, why not run an old miner instead of a conventional electric heater? One could even imagine die hard enthusiasts hooking up miners to thermostats and time switches to supply heat when needed.
What would be a decent 1kW miner to buy, as a space heater, I wonder? This time next year I imagine I will be able to pick up a 4-module Avalon for little more than the cost of a fan heater :-)
-roy
September 03, 2013, 05:05:05 AM
There's also lots of people with "free" electricity (as in, someone else pays for it). For instance my rent includes bill within a certain usage. I run ASICs and honestly I don't look at their consumption (well, I did just for curiosity, but I'm not tracking it).
This is rather common in Europe and Asia.
This is rather common in Europe and Asia.
I think you mean Eastern Europe and not Europe in general...
Quote
Obviously these are not going to be big operations, but many people can add up to be significant. Depends on how decentralised mining ends up being for the next few years.
Even with free power, some people ponied up more BTC for an ASIC (Batch #3 from Avalon for example) than it can ever generate due to the rise in difficulty. I guess some will be so stubborn to let the machines run to recover whatever they can.
I'm in London, UK. Lots of places include bills in London and elsewhere in the UK, as long as you don't exceed a certain "fair usage" limit. I've had similar experiences in Spain, France, Italy. Also in a number of countries in Asia. Only in the US, from the places that I've been, such options are rare.
So no, not just Eastern Europe.
There is nothing stubborn about letting your machines generate whatever they can when your electricity bill is the same. Unlike GPUs, you cannot really resell them or use them for something else.
September 03, 2013, 04:57:26 AM
I am scanning this thread and can't see the word "refund" anywhere.
I think if difficulty continues this 30% increase every 10 days, people will start canceling their pre orders and ask for refunds.
So, IMO, to get a more realistic prediction one should check all the refunds policies of all manufactures.
I think if difficulty continues this 30% increase every 10 days, people will start canceling their pre orders and ask for refunds.
So, IMO, to get a more realistic prediction one should check all the refunds policies of all manufactures.
September 01, 2013, 08:08:18 AM
I doubt even that will happen with ASICs. With GPUs, people had other uses for them. With ASICs, lots of people are not even aware of their consumption and they don't have any other usage for them. I don't think they will track the price so tightly if at all... but I guess we will eventually find out when we get closer to break-even.
I think it's highly unlikely that people will order custom hardware like ASICs to mine BTC and NOT track the power consumption and do the math (unless they are the kind of customers BFL got where Josh was laughing at them buying items where they had no clue what it did)
There's also lots of people with "free" electricity (as in, someone else pays for it). For instance my rent includes bill within a certain usage. I run ASICs and honestly I don't look at their consumption (well, I did just for curiosity, but I'm not tracking it).
This is rather common in Europe and Asia.
Obviously these are not going to be big operations, but many people can add up to be significant. Depends on how decentralised mining ends up being for the next few years.
Actually you gave me a new point of view on this: Maybe those "exploiting free energy" people will be the only ones keeping us from a centralized mining environment? Big projects will drive small projects out of business, but those "free energy" guys can only do small operations before they will have trouble..
Ente
September 01, 2013, 04:33:17 AM
You gave me a rational reason to be bullish on the bitcoin price.
September 01, 2013, 03:22:05 AM
There's also lots of people with "free" electricity (as in, someone else pays for it). For instance my rent includes bill within a certain usage. I run ASICs and honestly I don't look at their consumption (well, I did just for curiosity, but I'm not tracking it).
This is rather common in Europe and Asia.
This is rather common in Europe and Asia.
I think you mean Eastern Europe and not Europe in general...
Quote
Obviously these are not going to be big operations, but many people can add up to be significant. Depends on how decentralised mining ends up being for the next few years.
Even with free power, some people ponied up more BTC for an ASIC (Batch #3 from Avalon for example) than it can ever generate due to the rise in difficulty. I guess some will be so stubborn to let the machines run to recover whatever they can.
September 01, 2013, 02:27:02 AM
I doubt even that will happen with ASICs. With GPUs, people had other uses for them. With ASICs, lots of people are not even aware of their consumption and they don't have any other usage for them. I don't think they will track the price so tightly if at all... but I guess we will eventually find out when we get closer to break-even.
I think it's highly unlikely that people will order custom hardware like ASICs to mine BTC and NOT track the power consumption and do the math (unless they are the kind of customers BFL got where Josh was laughing at them buying items where they had no clue what it did)
There's also lots of people with "free" electricity (as in, someone else pays for it). For instance my rent includes bill within a certain usage. I run ASICs and honestly I don't look at their consumption (well, I did just for curiosity, but I'm not tracking it).
This is rather common in Europe and Asia.
Obviously these are not going to be big operations, but many people can add up to be significant. Depends on how decentralised mining ends up being for the next few years.
August 31, 2013, 07:04:04 PM
That is a good point. Another way to look at it is difficulty may overshoot before correcting. We say this multiple times over the years with GPUs. The trend is something like, exchange rate rises rapidly and then difficulty follows over a much slower timeframe as miners deploy new hardware prices, eventually prices peak and decline and thus pushes the most inefficient miners into a negative operating margin. Maybe they hang on for a while hoping difficulty goes lower or prices rise but eventually they give up and difficulty declines back towards equilibrium.
If the electrical cost to mine a Bitcoin is slightly higher than purchase cost miners may continue to mine for some time. One reason is that it is an easy anonymous way to "buy" Bitcoins from your power company. However it is important to keep in mind that unlike the GPU world the efficiency between the least efficient rigs and the most efficient ones is on the order of 12x (possibly more if/when Bitfury produces a 28nm device). So while electricity is still a small % of gross revenue for a cointerra miner it will be >150% of gross revenue for an Avalon one. So difficulty can continue to rise far beyond the break even point of an Avalon miner to a level where even the most "stubborn" miner will admit defeat (Paying $5 in electricity for $1 in BTC).
If the electrical cost to mine a Bitcoin is slightly higher than purchase cost miners may continue to mine for some time. One reason is that it is an easy anonymous way to "buy" Bitcoins from your power company. However it is important to keep in mind that unlike the GPU world the efficiency between the least efficient rigs and the most efficient ones is on the order of 12x (possibly more if/when Bitfury produces a 28nm device). So while electricity is still a small % of gross revenue for a cointerra miner it will be >150% of gross revenue for an Avalon one. So difficulty can continue to rise far beyond the break even point of an Avalon miner to a level where even the most "stubborn" miner will admit defeat (Paying $5 in electricity for $1 in BTC).
Yes, I agree with you on that.
But to continue, my feeling is that almost no one buying recently or now or in the near future will get ROI, both due to overoptimism and also due to a significant number of unprofitable miners hanging on until they finally hit their pain point - whereever that is.
I think we're looking at a period of time - maybe a year - where virtually no one will get ROI. And following that, ROI will only return when the prevailing wisdom on mining changes from bullish to bearish - i.e. when the majority of posts here are saying that mining is a mugs' game and that the only way you can make money is if you either have free electricity or are like ASICMINER designing your own chips.
What I haven't decided is whether that change in sentiment is going to happen gradually or suddenly. But I'm pretty sure sentiment will overshoot, and when we switch from the current prevailing wisdom that there is money to be made in mining, to one that there isn't - then, of course, the 'pessimism bias' will make it much easier for the miners who keep the faith to attain ROI.
I'm leaning towards 'fairly sudden', I think, but not sure... ETA: 'Sudden' is good of course, because it means more overshoot in sentiment...
roy
August 31, 2013, 06:44:12 PM
So, assuming that one is trying to accumulate as many BTC as possible gambling that it will go higher...hell, why not right?
You could subsidize an equal amount Gh/s of GPUs with say an equal amount of Gh/s of AsicMiner USBs or blades, and average it out until you hit 1.2-ish billion difficulty.
The ASICs would be subsidizing the power consumption of your GPUs (assuming they have long been paid for), and the rise in BTC price would cancel out some of your power use.
That might even be a net win over the long term?
Opinions?
You could subsidize an equal amount Gh/s of GPUs with say an equal amount of Gh/s of AsicMiner USBs or blades, and average it out until you hit 1.2-ish billion difficulty.
The ASICs would be subsidizing the power consumption of your GPUs (assuming they have long been paid for), and the rise in BTC price would cancel out some of your power use.
That might even be a net win over the long term?
Opinions?
August 31, 2013, 04:45:59 PM
The cost to reach the "break even point" is based on current cost so I don't find it too useful, because cost will decline significantly. The network will approach a small margin below the break even point.
I've always assumed that, in the short-to-medium term, the equilibrium point will be on the wrong side of breakeven, due to optimism bias in the purchase decisions. It's true it wouldn't make much financial sense to continue to mine under those circumstances, but there are other reasons to hash than profit, and even the people who were in it purely for the money will find other ways to justify their purchases to themselves.
roy
That is a good point. Another way to look at it is difficulty may overshoot before correcting. We say this multiple times over the years with GPUs. The trend is something like, exchange rate rises rapidly and then difficulty follows over a much slower timeframe as miners deploy new hardware prices, eventually prices peak and decline and thus pushes the most inefficient miners into a negative operating margin. Maybe they hang on for a while hoping difficulty goes lower or prices rise but eventually they give up and difficulty declines back towards equilibrium.
If the electrical cost to mine a Bitcoin is slightly higher than purchase cost miners may continue to mine for some time. One reason is that it is an easy anonymous way to "buy" Bitcoins from your power company. However it is important to keep in mind that unlike the GPU world the efficiency between the least efficient rigs and the most efficient ones is on the order of 12x (possibly more if/when Bitfury produces a 28nm device). So while electricity is still a small % of gross revenue for a cointerra miner it will be >150% of gross revenue for an Avalon one. So difficulty can continue to rise far beyond the break even point of an Avalon miner to a level where even the most "stubborn" miner will admit defeat (Paying $5 in electricity for $1 in BTC).
August 31, 2013, 04:12:46 PM
The cost to reach the "break even point" is based on current cost so I don't find it too useful, because cost will decline significantly. The network will approach a small margin below the break even point.
I've always assumed that, in the short-to-medium term, the equilibrium point will be on the wrong side of breakeven, due to optimism bias in the purchase decisions. It's true it wouldn't make much financial sense to continue to mine under those circumstances, but there are other reasons to hash than profit, and even the people who were in it purely for the money will find other ways to justify their purchases to themselves.
roy
I agree.
They may have put a lot of money into the hardware, maybe even more than they should have afforded. With no money left, they simply can't shut down mining and buy bitcoins directly. So it could be "exchanging electricity-money to bitcoins".
Expecting the price to rise is easy when you are a bull anyway and invested in hardware. So this dip is just temporary.
And finally, noone likes to accept to have made a mistake.
..not even talking about parents paying for electricity, or the landlord splitting the bill among several flats..
Ente
August 31, 2013, 02:15:04 PM
The cost to reach the "break even point" is based on current cost so I don't find it too useful, because cost will decline significantly. The network will approach a small margin below the break even point.
I've always assumed that, in the short-to-medium term, the equilibrium point will be on the wrong side of breakeven, due to optimism bias in the purchase decisions. It's true it wouldn't make much financial sense to continue to mine under those circumstances, but there are other reasons to hash than profit, and even the people who were in it purely for the money will find other ways to justify their purchases to themselves.
roy
August 30, 2013, 11:07:48 AM
I doubt even that will happen with ASICs. With GPUs, people had other uses for them. With ASICs, lots of people are not even aware of their consumption and they don't have any other usage for them. I don't think they will track the price so tightly if at all... but I guess we will eventually find out when we get closer to break-even.
I think it's highly unlikely that people will order custom hardware like ASICs to mine BTC and NOT track the power consumption and do the math (unless they are the kind of customers BFL got where Josh was laughing at them buying items where they had no clue what it did)
August 30, 2013, 08:58:16 AM
@ muyuu,
I think you misread (or I wasn't clear). The scenario is the network is in equilibrium (hashrate growth changes with exchange rate because the difficulty is near the collective break even point for the network). To illustrate lets say the exchange rate is $200 per BTC and there is a miner with an older 130nm rig which requires $180 in electricity to produce 1 BTC (based on difficulty, efficiency, and power cost). Now the exchange rate falls and remains below $150 for thirty days. At some point that miner is going to stop producing one BTC worth <$150 using >$180 in electricity and just buy BTC because it is cheaper. People may make bad long term plans but generally make better short term ones. When you are spending >1 BTC to produce 1 BTC it is pretty easy to see you should stop. On the other hand he is unlikely to throw his rig away. If the price (or more importantly price/difficulty) rebounds then we will turn his rig on. The bad news for this marginal miner is that the higher price/difficult will bring in new investment likely by miners with lower power cost and more efficient gear so difficulty will likely overshoot his break even point and he will have to idle again.
We aren't concerned with hardware costs for the break even point as you indicate they are a sunk cost even if a miner has a lifetime loss it still makes sense to mine to reduce that loss UNTIL difficulty exceeds the break even point. At that point any additional mining is just increasing the loss for the miner.
We saw this in dynamic in the GPU era where difficulty would track price. Today it doesn't simply because we are no where near the break even point for the network. For even the least efficient gear and the highest power cost user it would be >6PH/s.
I think you misread (or I wasn't clear). The scenario is the network is in equilibrium (hashrate growth changes with exchange rate because the difficulty is near the collective break even point for the network). To illustrate lets say the exchange rate is $200 per BTC and there is a miner with an older 130nm rig which requires $180 in electricity to produce 1 BTC (based on difficulty, efficiency, and power cost). Now the exchange rate falls and remains below $150 for thirty days. At some point that miner is going to stop producing one BTC worth <$150 using >$180 in electricity and just buy BTC because it is cheaper. People may make bad long term plans but generally make better short term ones. When you are spending >1 BTC to produce 1 BTC it is pretty easy to see you should stop. On the other hand he is unlikely to throw his rig away. If the price (or more importantly price/difficulty) rebounds then we will turn his rig on. The bad news for this marginal miner is that the higher price/difficult will bring in new investment likely by miners with lower power cost and more efficient gear so difficulty will likely overshoot his break even point and he will have to idle again.
We aren't concerned with hardware costs for the break even point as you indicate they are a sunk cost even if a miner has a lifetime loss it still makes sense to mine to reduce that loss UNTIL difficulty exceeds the break even point. At that point any additional mining is just increasing the loss for the miner.
We saw this in dynamic in the GPU era where difficulty would track price. Today it doesn't simply because we are no where near the break even point for the network. For even the least efficient gear and the highest power cost user it would be >6PH/s.
I doubt even that will happen with ASICs. With GPUs, people had other uses for them. With ASICs, lots of people are not even aware of their consumption and they don't have any other usage for them. I don't think they will track the price so tightly if at all... but I guess we will eventually find out when we get closer to break-even.
August 29, 2013, 08:37:03 PM
Good list.
Bitfury is somewhat better than that though, I measured +116W for adding a 150GH/s Bitfury unit to my rig. Gold-rated 1300W PSU at 440W@230V total now so I'd assume about 90% efficiency. Anyway, <0.8J/GH at the wall for Bitfury. Another guy had similar results.
Bitfury is somewhat better than that though, I measured +116W for adding a 150GH/s Bitfury unit to my rig. Gold-rated 1300W PSU at 440W@230V total now so I'd assume about 90% efficiency. Anyway, <0.8J/GH at the wall for Bitfury. Another guy had similar results.
Was this on the reduced speed boards? The reason I ask is if the boards intended for 25 GH/s are getting 21 GH/s then they are essentially underclocked (even if unwillingly) and that will improve efficiency.
Was the 116W increase at the wall?
Yes on both counts. And right, I suppose they might be somewhat less efficient at full speeds.
August 29, 2013, 07:12:14 PM
Cointerra has the most power efficient 28nm process at the moment.
1666 MH/J
1666 MH/J
Do you have a link to this. I thought Cointerra simply said significantly below 1GH/W. Also I need full system power not the chip. Even if it is an estimate the manufacturer I will use it if they are willing to put it into writing.
You can find it here:
http://thegenesisblock.com/cointerra-announces-2ths-asic-bitcoin-miner-for-15750/
Thanks. I sent a PM to cointerra to confirm. The fact that it is only on coindesk and not repeated anywhere on Cointerra website or in any of their posts makes me want confirmation. I asked them also to confirm is the wattage estimate is at the chip level or at the wall.
For BFL I likely will need to estimate an average system load to make it apples to apples. That is ugly but with every other ASIC manufacturer building complete systems I don't see a better way.
Another thing to consider is that BFL claims you will also have the option to use USB interface instead of pci-e as a standalone device as well. I'll try and dig you up a link....
Edit: Here's info (Monarch btw)
https://products.butterflylabs.com/homepage-new-products/600-gh-bitcoin-mining-card.html
Quote
Connectivity
USB 2.0 - Monarch cards can be used as an external computer peripherial and chained via USB hub. In this mode it can be controlled via an Android host or standard Linux or Windows computer.
PCI-Express - Monarch cards consume two PCI slots when installed in a standard ATX montherboard. The PCIe format used is x1 for maximum compatibility.
USB 2.0 - Monarch cards can be used as an external computer peripherial and chained via USB hub. In this mode it can be controlled via an Android host or standard Linux or Windows computer.
PCI-Express - Monarch cards consume two PCI slots when installed in a standard ATX montherboard. The PCIe format used is x1 for maximum compatibility.
August 29, 2013, 06:58:40 PM
Good list.
Bitfury is somewhat better than that though, I measured +116W for adding a 150GH/s Bitfury unit to my rig. Gold-rated 1300W PSU at 440W@230V total now so I'd assume about 90% efficiency. Anyway, <0.8J/GH at the wall for Bitfury. Another guy had similar results.
Bitfury is somewhat better than that though, I measured +116W for adding a 150GH/s Bitfury unit to my rig. Gold-rated 1300W PSU at 440W@230V total now so I'd assume about 90% efficiency. Anyway, <0.8J/GH at the wall for Bitfury. Another guy had similar results.
Was this on the reduced speed boards? The reason I ask is if the boards intended for 25 GH/s are getting 21 GH/s then they are essentially underclocked (even if unwillingly) and that will improve efficiency.
Was the 116W increase at the wall?
August 29, 2013, 06:49:31 PM
Good list.
Bitfury is somewhat better than that though, I measured +116W for adding a 150GH/s Bitfury unit to my rig. Gold-rated 1300W PSU at 440W@230V total now so I'd assume about 90% efficiency. Anyway, <0.8J/GH at the wall for Bitfury. Another guy had similar results.
Bitfury is somewhat better than that though, I measured +116W for adding a 150GH/s Bitfury unit to my rig. Gold-rated 1300W PSU at 440W@230V total now so I'd assume about 90% efficiency. Anyway, <0.8J/GH at the wall for Bitfury. Another guy had similar results.
August 29, 2013, 05:58:13 PM
Lowered efficiency of Avalon from 8.5 J/GH to 8.8 J/GH. The Avalon miner is shipped in multiple configurations with 2, 3, or 4 modules in three different batches, using different power supplies. The goal of the table is to provide an average break even point. Reported efficiency "at the wall" ranges from 8.5 J/GH to 9.3 J/GH.
August 29, 2013, 02:57:04 PM
.0718 kwh here in Manitoba and lots of cold air in the winter
August 29, 2013, 02:20:44 PM
Thanks for this OP. I've had this conversation so many times but needed the numbers to back it up!
August 29, 2013, 01:22:33 PM
@ muyuu,
I think you misread (or I wasn't clear). The scenario is the network is in equilibrium (hashrate growth changes with exchange rate because the difficulty is near the collective break even point for the network). To illustrate lets say the exchange rate is $200 per BTC and there is a miner with an older 130nm rig which requires $180 in electricity to produce 1 BTC (based on difficulty, efficiency, and power cost). Now the exchange rate falls and remains below $150 for thirty days. At some point that miner is going to stop producing one BTC worth <$150 using >$180 in electricity and just buy BTC because it is cheaper. People may make bad long term plans but generally make better short term ones. When you are spending >1 BTC to produce 1 BTC it is pretty easy to see you should stop. On the other hand he is unlikely to throw his rig away. If the price (or more importantly price/difficulty) rebounds then we will turn his rig on. The bad news for this marginal miner is that the higher price/difficult will bring in new investment likely by miners with lower power cost and more efficient gear so difficulty will likely overshoot his break even point and he will have to idle again.
We aren't concerned with hardware costs for the break even point as you indicate they are a sunk cost even if a miner has a lifetime loss it still makes sense to mine to reduce that loss UNTIL difficulty exceeds the break even point. At that point any additional mining is just increasing the loss for the miner.
We saw this in dynamic in the GPU era where difficulty would track price. Today it doesn't simply because we are no where near the break even point for the network. For even the least efficient gear and the highest power cost user it would be >6PH/s.
I think you misread (or I wasn't clear). The scenario is the network is in equilibrium (hashrate growth changes with exchange rate because the difficulty is near the collective break even point for the network). To illustrate lets say the exchange rate is $200 per BTC and there is a miner with an older 130nm rig which requires $180 in electricity to produce 1 BTC (based on difficulty, efficiency, and power cost). Now the exchange rate falls and remains below $150 for thirty days. At some point that miner is going to stop producing one BTC worth <$150 using >$180 in electricity and just buy BTC because it is cheaper. People may make bad long term plans but generally make better short term ones. When you are spending >1 BTC to produce 1 BTC it is pretty easy to see you should stop. On the other hand he is unlikely to throw his rig away. If the price (or more importantly price/difficulty) rebounds then we will turn his rig on. The bad news for this marginal miner is that the higher price/difficult will bring in new investment likely by miners with lower power cost and more efficient gear so difficulty will likely overshoot his break even point and he will have to idle again.
We aren't concerned with hardware costs for the break even point as you indicate they are a sunk cost even if a miner has a lifetime loss it still makes sense to mine to reduce that loss UNTIL difficulty exceeds the break even point. At that point any additional mining is just increasing the loss for the miner.
We saw this in dynamic in the GPU era where difficulty would track price. Today it doesn't simply because we are no where near the break even point for the network. For even the least efficient gear and the highest power cost user it would be >6PH/s.
August 29, 2013, 09:04:12 AM
So it looks pretty safe to say that the network will be less than 250 PH/s using the 28nm technology at the current prices.
The replacement cost is calculated based on all hashing power coming from a single design so it shouldn't be taken as likely but more an upper bound. I think there might be an error in the math as well.
The cost to reach the "break even point" is based on current cost so I don't find it too useful, because cost will decline significantly. The network will approach a small margin below the break even point. It is only a matter of time. You can consider that the equilibrium point, we have seen it occur in the GPU era for a couple of years. If more efficient hardware is produced then the hashrate will rise, if the exchange rate rises then the hashrate will rise, as mining moves to lower cost areas the hashrate will rise, if the exchange rate falls the hashrate will fall. However the network will approach and stay close to a margin below the break even point. How close? It really depends on miners and what risk they are taking.
I'm not so sure of the conclusion that a price fall will be closely followed by a hashrate fall. ASICs have mostly no alternative use and there is no outside market for them.
Provided they are in profit, they will remain on. Even at a loss many people won't bother turning them off.
August 28, 2013, 06:54:54 PM
Quote from: KNC
Power Supply Recommendation.
With our shipment date approaching this update addresses the power demands of our mining devices.
We realise we are currently being compared to our competitors with respect to power consumption and would like to clarify our position.
Today we can reveal that our maximal power consumption will be below 1.6 W/GH/s.
We understand the need for some of our more remote customers to be able to secure a purchase of a power supply (PSU) within the given timeframe.
So today we can reveal the following recommendations;
We recommend an 850 Watt PSU with a minimum of 4x PCI-E adaptors for our Jupiter model.
We recommend a 600 Watt PSU with a minimum of 2x PCI-E adaptors for our Saturn model.
We recommend a 400 Watt PSU with a minimum of a PCI-E adaptor for our Mercury model.
This recommended power wattage figure is calculated upon our max. power consumption of total device including all of its components.
We aim to ensure you use a power supply unit capable of outputting in excess of the current recommended wattage to prevent any problems caused due to insufficient power.
Note: Powers supplies must contain a sufficient number of PCI-E adaptors for each respective unit.
https://www.kncminer.com/news/news-31With our shipment date approaching this update addresses the power demands of our mining devices.
We realise we are currently being compared to our competitors with respect to power consumption and would like to clarify our position.
Today we can reveal that our maximal power consumption will be below 1.6 W/GH/s.
We understand the need for some of our more remote customers to be able to secure a purchase of a power supply (PSU) within the given timeframe.
So today we can reveal the following recommendations;
We recommend an 850 Watt PSU with a minimum of 4x PCI-E adaptors for our Jupiter model.
We recommend a 600 Watt PSU with a minimum of 2x PCI-E adaptors for our Saturn model.
We recommend a 400 Watt PSU with a minimum of a PCI-E adaptor for our Mercury model.
This recommended power wattage figure is calculated upon our max. power consumption of total device including all of its components.
We aim to ensure you use a power supply unit capable of outputting in excess of the current recommended wattage to prevent any problems caused due to insufficient power.
Note: Powers supplies must contain a sufficient number of PCI-E adaptors for each respective unit.
Raised (improved) estimate for KNC to 1.8 J/GH (from 2.5 J/GH) at the wall. KNC reports efficiency is less than 1.6 J/GH however it is unclear what the reported efficiency includes. I have asked KNC for a clarification but they haven't responded yet. The break even difficulty is based on wattage "at the wall" (120V/230V measured/estimated at the plug).
If 1.6 J/GH refers to the ...
... system efficiency at the wall (AC load) then efficiency in table should be 1.6 J/GH.
... system total DC load then efficiency in the table should be 1.8 J/GH.
... ASIC boards (chips + DC PSU) but excludes the fans, controller, and ATX PSU then efficiency in the table should be 1.9 J/GH.
... ASIC chips and excludes the DC to DC supply, fans, controllers, and ATX PSU then efficiency in the table should be 2.0 J/GH.
I feel that based on the wording of the news and considering the PSU is not included, the second scenario is the most likely so the efficiency table will reflect 1.8 J/GH until more information is available. I would have imagined that if KNC was reported "at the wall" efficiency they would have conditioned it with language like "with 80Plus Gold or better power supply, "as tested on power supply model #####", or "assuming 90% DC to AC efficiency".
On edit: KNC supporters you can stop PM me. I will change the table when either KNC confirms the "at the wall" efficiency or we have user provided benchmarks in the field. If you want it updated, ask KNC to provide clarification.
August 27, 2013, 09:21:50 PM
mark！
August 27, 2013, 04:41:35 PM
Updated to include estimates for Cointerra & Monarch. Given the high efficiency of most devices switched efficiency units to J/GH.
August 26, 2013, 02:15:05 PM
Cointerra has the most power efficient 28nm process at the moment.
1666 MH/J
1666 MH/J
Do you have a link to this. I thought Cointerra simply said significantly below 1GH/W. Also I need full system power not the chip. Even if it is an estimate the manufacturer I will use it if they are willing to put it into writing.
You can find it here:
http://thegenesisblock.com/cointerra-announces-2ths-asic-bitcoin-miner-for-15750/
Thanks. I sent a PM to cointerra to confirm. The fact that it is only on coindesk and not repeated anywhere on Cointerra website or in any of their posts makes me want confirmation. I asked them also to confirm is the wattage estimate is at the chip level or at the wall.
For BFL I likely will need to estimate an average system load to make it apples to apples. That is ugly but with every other ASIC manufacturer building complete systems I don't see a better way.
August 26, 2013, 01:56:12 PM
So it looks pretty safe to say that the network will be less than 250 PH/s using the 28nm technology at the current prices.
The replacement cost is calculated based on all hashing power coming from a single design so it shouldn't be taken as likely but more an upper bound. I think there might be an error in the math as well.
The cost to reach the "break even point" is based on current cost so I don't find it too useful, because cost will decline significantly. The network will approach a small margin below the break even point. It is only a matter of time. You can consider that the equilibrium point, we have seen it occur in the GPU era for a couple of years. If more efficient hardware is produced then the hashrate will rise, if the exchange rate rises then the hashrate will rise, as mining moves to lower cost areas the hashrate will rise, if the exchange rate falls the hashrate will fall. However the network will approach and stay close to a margin below the break even point. How close? It really depends on miners and what risk they are taking.
August 26, 2013, 01:05:17 PM
So it looks pretty safe to say that the network will be less than 250 PH/s using the 28nm technology at the current prices.
August 26, 2013, 11:46:07 AM
Cointerra has the most power efficient 28nm process at the moment.
1666 MH/J
1666 MH/J
Do you have a link to this. I thought Cointerra simply said significantly below 1GH/W. Also I need full system power not the chip. Even if it is an estimate the manufacturer I will use it if they are willing to put it into writing.
You can find it here:
http://thegenesisblock.com/cointerra-announces-2ths-asic-bitcoin-miner-for-15750/
August 26, 2013, 11:39:19 AM
Code:
Specs Process Eff (MH/J) Diff (mil) Hashrate (TH/s) Cost(mil)
BFL 28nm 1666 34,925 250,000 1,950
Cointerra 28nm 1666 34,925 250,000 1,969
Hashfast 28nm 1200 25,146 180,000 2,520
KNC 28nm 400 8,382 60,000 1,050
Device Process Eff (MH/J) Diff (mil) Hashrate (TH/s)
Avalon 130nm 120 2,515 18,000
ASICMiner 110nm 130 2,724 19,500
BFL 65nm 200 4,191 30,000
Bitfury 55nm 1000 20,955 150,000
FPGA various 20 419 3,000
GPU various 3 63 450
The issue here is you are comparing the chip level for one product to the board level for another product to the system level for the rest of the products.
For example if BFL delivers a 600 GH/s card which has 1,666 MH/J efficiency it requires a host system with one or more PCIe slots (50W min), case fans (3x12W ea?), and a AC PSU which is say 90% efficient at load. So BFL would have a total DC load of 436W which is a 484W AC load (assume 90% efficient PSU). 600 GH/s / 484W = 1240 MH/J. Of course with more fans, higher system wattage, or worse PSU it would be even lower.
Cointerra is even more an unknown. We don't even know the board level wattage (@12V), what type of controller will be used, is it a custom enclosure, if so how many fans, etc. That makes trying to estimate a system wattage totally unknown. Also as indicated above I can't see to find a cite for Cointerra indicating a specific efficiency even just for the chip. However even if Cointerra chip(s) deliver 2TH/s using 1200W the ASIC (like all other ASICs will run at @0.8V to 1.2V) and need a DC to DC PSU to convert the 12V supplied by the AC PSU to the ~1V required by the chip. The DC wattage would then be 1200/0.9 = 1,333W @ 12VDC. If we assume the balance of the system is a Pi or other micro controller (5W and 3x 1A fans) it would make total system wattage 1374W DC = 1527W AC. That would put the efficiency at the plug ~1300 MH/J.
Simple version
System AC ("at the wall") wattage = System DC Wattage / ATX PSU efficiency *
System DC wattage = ASIC Board(s) wattage + balance of system (controllers, fans, waterpump, etc) **
Board DC Wattage = ASIC Wattage / DC to DC PSU ***
* Depends on ATX PSU efficiency at the particular load. As a power customer you pay for wattage at the wall. This will always be higher than the DC wattage used by the system due to PSU inefficiency. If unknown a good upper bound is 90% efficiency. More efficient PSU are possible but they generally require 230V and are more expensive (google "80 Plus Titanium")
** Rasberry Pi uses <5W. BFL PCIe card will require a system with one or more PCIe slots which means a "traditional" motherboard and >50W. If case fans or pumps are unknown an estimate of 12W per 120mm fan and 20W per pump (HashFast) is reasonable. Without knowing the exact components we can't know the exact system wattage but for example nobody is going to use fans which use half a watt or a water pump that uses 100W or PC host which uses 10W.
*** ATX PSU supplies high current power at 12V however no ASIC runs at that voltage. Each ASIC board will need to convert the 12V to the voltage used by the ASIC. This will vary from ASIC to ASIC but is generally in the range of 0.6V to 1.3V. Good high current DC to DC power supplies are expensive (up to $1 per watt) and generally are less than 90% efficient. Lower cost PSU will provide significantly reduced efficiency (80% to 84%).
August 26, 2013, 11:29:27 AM
Cointerra has the most power efficient 28nm process at the moment.
1666 MH/J
1666 MH/J
Do you have a link to this. I thought Cointerra simply said significantly below 1GH/W. Also I need full system power not the chip. Even if it is an estimate the manufacturer I will use it if they are willing to put it into writing.
August 26, 2013, 09:37:53 AM
Here I've added cost of the cheapest technology to make this particular technology obsolete
Code:
Specs Process Eff (MH/J) Diff (mil) Hashrate (TH/s) Cost(mil) Minimum Replacement Cost(mil)
BFL 28nm 1666 34,925 250,000 1,950 1,950
Cointerra 28nm 1666 34,925 250,000 1,969 1,950
Hashfast 28nm 1200 25,146 180,000 2,520 1,440
KNC 28nm 400 8,382 60,000 1,050 480
Device Process Eff (MH/J) Diff (mil) Hashrate (TH/s) Cost(mil)
Avalon 130nm 120 2,515 18,000 2,520 144
ASICMiner 110nm 130 2,724 19,500 1,872 156
BFL 65nm 200 4,191 30,000 1,410 240
Bitfury 55nm 1000 20,955 150,000 3,000 1,200
FPGA various 20 419 3,000
GPU various 3 63 450
August 26, 2013, 08:39:16 AM
Code:
Specs Process Eff (MH/J) Diff (mil) Hashrate (TH/s) Cost(mil)
BFL 28nm 1666 34,925 250,000 1,950
Cointerra 28nm 1666 34,925 250,000 1,969
Hashfast 28nm 1200 25,146 180,000 2,520
KNC 28nm 400 8,382 60,000 1,050
Device Process Eff (MH/J) Diff (mil) Hashrate (TH/s)
Avalon 130nm 120 2,515 18,000
ASICMiner 110nm 130 2,724 19,500
BFL 65nm 200 4,191 30,000
Bitfury 55nm 1000 20,955 150,000
FPGA various 20 419 3,000
GPU various 3 63 450
August 26, 2013, 07:55:22 AM
Anybody care to add a calculation for the total cost to achieve the listed hashrate for each of the respective models?
ie.
for the HashFast product and corresponding breakeven point of 25trillion it's $352Million in hardware costs
For BFL's 65nm is $188Million
etc.
ie.
for the HashFast product and corresponding breakeven point of 25trillion it's $352Million in hardware costs
For BFL's 65nm is $188Million
etc.
August 26, 2013, 07:38:15 AM
It certainly puts a feather in the cap for Bitfury. Of the shipping chips they are way out ahead of the pack (by 5x!).
Though at current pricing of $35 -> $54 per Gigahash (depending on overclock), the pricing is a bit meh.
https://megabigpower.com/shop/index.php?route=product/product&path=60&product_id=62
Though at current pricing of $35 -> $54 per Gigahash (depending on overclock), the pricing is a bit meh.
https://megabigpower.com/shop/index.php?route=product/product&path=60&product_id=62
August 26, 2013, 07:22:58 AM
I believe this thread is trying to calculate the "theoretical maximum economically sustainable system hash rate given various technologies", right?
Therefore all vaporware should simply be included in a separate category:
BTW, Thanks! And this is very interesting.
Therefore all vaporware should simply be included in a separate category:
Code:
Specs Process Eff (MH/J) Diff (mil) Hashrate (TH/s)
BFL 28nm ...
Cointerra 28nm ...
Hashfast 28nm 1200 25,146 180,000
KNC 28nm 400 8,382 60,000
Device Process Eff (MH/J) Diff (mil) Hashrate (TH/s)
Avalon 130nm 120 2,515 18,000
ASICMiner 110nm 130 2,724 19,500
BFL 65nm 200 4,191 30,000
Bitfury 55nm 1000 20,955 150,000
FPGA various 20 419 3,000
GPU various 3 63 450
BTW, Thanks! And this is very interesting.
August 26, 2013, 07:09:22 AM
Really?
They demonstrated that? It's for whole platform (like in OP charts) or chip itself? From 1666 you may easily get down to 1200 with poor PSU design...
I guess with KNC and Hashfast also in the list, you have to be consistent and also include Cointerra.
OR
You exclude all chips currently under development.
August 26, 2013, 07:03:38 AM
Cointerra has the most power efficient 28nm process at the moment.
1666 MH/J
Really?1666 MH/J
They demonstrated that? It's for whole platform (like in OP charts) or chip itself? From 1666 you may easily get down to 1200 with poor PSU design...
Most efficient vaporware indeed
I hate it when people start stating claims based on marketing materials as if they are a fact...
August 26, 2013, 06:57:10 AM
Cointerra has the most power efficient 28nm process at the moment.
1666 MH/J
Really?1666 MH/J
They demonstrated that? It's for whole platform (like in OP charts) or chip itself? From 1666 you may easily get down to 1200 with poor PSU design...
August 26, 2013, 06:03:36 AM
Cointerra has the most power efficient 28nm process at the moment.
1666 MH/J
1666 MH/J
August 26, 2013, 05:34:00 AM
Do I read this correctly, GPU miners are already not getting enough btc to break even?
At $100 exchange rate and $0.10 electrical rate? That is correct. Obviously each miner's efficiency will vary as will their electrical rates and their "pain threshold" (how negative of a margin they are willing to accept before shutting down) but generally speaking yes.
August 26, 2013, 02:32:14 AM
Do I read this correctly, GPU miners are already not getting enough btc to break even?
August 25, 2013, 08:20:09 PM
$0.30+ per kWh reporting in 
Ouch. The break even difficulty and hashrate would be 1/3rd of what is listed.
August 25, 2013, 08:19:53 PM
$0.30+ per kWh reporting in
Hello fellow Californian?
August 25, 2013, 08:18:53 PM
$0.30+ per kWh reporting in
August 25, 2013, 07:33:38 PM
It is useful to know what the electrical break even difficulty of a device in estimating future hashrate growth of the network. A miner is unlikely to continue to mine for any extended period of time at a negative operating margin. The break even difficulty is the point where the electricity consumed by the mining equipment equals the value of the coins mined. At the break even point it will cost the miner the same amount to mine a coin as it would to purchase it. The hardware cost is not considered in the electrical break even point as once purchased the hardware is a sunk cost and the miner is very likely to continue mining as long as it shows a positive operating margin (electricity < value of BTC). Just because the difficulty is below the break even point doesn't guarantee the miner a positive return.
The electrical break even point is based on three factors:
a) the current exchange rate (USD per BTC)
b) the price the miner pays for power (USD per kWh)
c) the efficiency of the mining hardware (J/GH)https://bitcointalk.org/Themes/custom1/images/smflogo.gif
Electrical break even point by hardware efficiency
Based on $100 BTC:USD exchange rate and $0.10 per kWh electrical power rate
Update (05/25/2014): I no longer have the time to research every new ASIC device. I have not seen a device that is significantly more efficient than 0.8 J/GH (at full clock) when measured at the wall and multiple devices have real world efficiency within +/-20% of that spec so it is a good place to start estimating difficulty. Remember miners are paying for full power consumption not just the power of the raw chip. The at the wall measurement is the only thing that matters. More efficient designs may eventually emerge but the low hanging fruit is now gone. If less than a year efficient of best device went from >8 J/GH to <0.8 J/GH. We aren't going to see a 1000% improvement in efficient in the next year. Even a 0.4 J/GH device (at the wall) would be an impressive improvement but only represents a doubling of the break even point.
To calculate the break even difficulty at a different exchange and/or power rate use the following formula:
Examples:
Why J not W?
Because J (Joule) is a measure of energy and W (Watt) is a measure of power. One Joule per second is one watt. You can't compare watts (energy over time) to a timeless value like Gigahash. Efficiency can be measured either in W per GH/s or J/GH bu not W/GH. It would be like saying a car gets 30 MPH per gallon or the spot price of gold is $1500 per ounce second. You can safely assume is someone writes W/GH they mean J/GH (or W per GH/s). I put this here because I have already gotten "corrections" by PM. Simple version: W per GH/s or J/GH are correct, W/GH is not.
Why did you make this?
I find it silly people have projections with difficulty going to 1 trillion or more (I think the highest I have seen is 200 trillion). At a mere 300 billion difficulty all current and proposed mining devices would be operated at a significant loss (assuming $100 exchange rate & $0.10 electrical rate). So at 1 trillion in difficulty even a Cointera rig would be converting $4 in electricity into $1 in Bitcoins. Anyone think that is likely? Significantly higher difficulty is going to require either more efficient hardware, a higher exchange rate, or the average cost of power for the network to decline.
Notes:
[1] Numbers used are from mining hardware comparison where available or based on benchmarks reported by owners/manufacturer. For apples to apples comparison all units are based on wattage at the wall (including all system power and inefficiencies). For devices where total wattage at the wall is unavailable it has been estimated. For unreleased devices ("specs") the efficiency reported by manufacturer is used.
[2] Cointerra did not provide system efficiency "at the wall" however did indicate that chip wattage would be better than 0.55 J/GH, based on that I project the at wall efficiency to be ~0.7 J/GH. Updated to 0.6 J/GH at the chip and 0.75 J/GH at the wall based on official statement. This is based on the following assumptions. Chip power: 0.60 W per GH/s * 2000 GH/s = 1200W @ 0.785 VDC. DC to DC board PSU 90% efficient 1200W/0.9 = 1333W @ 12VDC. Balance of system controller & cooling/fans = 50W @ 12 VDC. Total system DC wattage 1383W @ 12VDC. ATX PSU 90% efficient 13832/0.9 = 1537W @ 120VAC. 1537W / 2000 GH/s = ~0.8 J/GH.
[3] Updated to 1.1 J/GH (https://bitcointalksearch.org/topic/m.3307091). 455W @ 1 VDC running at 502 GH/s. Assume 90% DC to DC conversion and 90% AC power supply efficiency results in estimated 560W at the wall.
[4] The Monarch card is reported by BFL to use 350W. It is not clearly indicated if this is 350W at the chip or the card but I will assume the 350W is the total 12VDC wattage for entire card (to include one or more ASIC processors, host interface, card fan, and DC to DC board PSU. The balance of system is provided by the user so some assumptions are necessary to estimate efficiency at the wall. To be used in a PCIe slot the card requires a "traditional motherboard" with one or more PCIe slot, 100W min is assumed for system and fans. Total DC wattage for single card system is 450W DC. Assuming a 90% efficient ATX PSU 450W DC/0.9 = 500W @ 120V AC. 500W/600GH/s = ~0.8 J/GH. Actual wattage may be significantly higher depending on the components selected by the user.
[5] Given BFL's delivery track record and after the fact changes I don't feel comfortable making any prediction. I would recommend users be very conservative and pad the delivery timeframes provided by BFL.
[6] Due to multiple configurations power efficiency varies by number of modules, batch version, and PSU efficiency. The following efficiencies have been reported 9.4 J/GH (66 GH/s @ 620W), 8.8 J/GH (85 GH/s @ 750W), 8.5 J/GH (82 GH/s @ 700W).
[7] Based on reported efficiency at the wall of less than 350W and nominal performance of greater than 400 GH/s of the BabyJet with single hashing board.
The electrical break even point is based on three factors:
a) the current exchange rate (USD per BTC)
b) the price the miner pays for power (USD per kWh)
c) the efficiency of the mining hardware (J/GH)https://bitcointalk.org/Themes/custom1/images/smflogo.gif
Electrical break even point by hardware efficiency
Based on $100 BTC:USD exchange rate and $0.10 per kWh electrical power rate
Code:
Device Process Eff (J/GH) [1] Diff (mil) Hashrate (PH/s)
--------------------------------------------------------------------------------------------------
GPU various 330.0 60 0.4
FPGA various 50.0 400 2.9
Avalon 110nm 8.8 [6] 2,400 17.2
ASICMiner 130nm 7.7 2,700 19.3
BFL (SC) 65nm 5.0 4,200 30.1
KNC 28nm 1.1 [3] 18,250 130.6
Bitfury 55nm 0.9 23,300 166.8
Hashfast 28nm ~0.8 [7] 25,100 179.7
Cointerra 28nm ~0.8 [2] 25,100 179.7
BFL (Monarch) 28nm ~0.8 [4] 25,100 179.7
Update (05/25/2014): I no longer have the time to research every new ASIC device. I have not seen a device that is significantly more efficient than 0.8 J/GH (at full clock) when measured at the wall and multiple devices have real world efficiency within +/-20% of that spec so it is a good place to start estimating difficulty. Remember miners are paying for full power consumption not just the power of the raw chip. The at the wall measurement is the only thing that matters. More efficient designs may eventually emerge but the low hanging fruit is now gone. If less than a year efficient of best device went from >8 J/GH to <0.8 J/GH. We aren't going to see a 1000% improvement in efficient in the next year. Even a 0.4 J/GH device (at the wall) would be an impressive improvement but only represents a doubling of the break even point.
To calculate the break even difficulty at a different exchange and/or power rate use the following formula:
Code:
Adjusted difficulty = base difficulty * (exchange rate / $100) * ($0.10/power rate)
Examples:
Code:
Avalon break even difficulty @ $200 exchange rate and same power rate. 2,400 * ($200/$100) * ($0.10/$0.10) = 4,800
BFL (SC) break even difficulty @ $0.15 power rate and same exchange rate. 4,200 * ($100/$100) * ($0.10/$0.15) = 2,800
Bitfury break even difficulty @ $200 exchange rate and $0.05 power rate. 21,000 * ($200/$100) * ($0.10/$0.05) = 84,000
Why J not W?
Because J (Joule) is a measure of energy and W (Watt) is a measure of power. One Joule per second is one watt. You can't compare watts (energy over time) to a timeless value like Gigahash. Efficiency can be measured either in W per GH/s or J/GH bu not W/GH. It would be like saying a car gets 30 MPH per gallon or the spot price of gold is $1500 per ounce second. You can safely assume is someone writes W/GH they mean J/GH (or W per GH/s). I put this here because I have already gotten "corrections" by PM. Simple version: W per GH/s or J/GH are correct, W/GH is not.
Why did you make this?
I find it silly people have projections with difficulty going to 1 trillion or more (I think the highest I have seen is 200 trillion). At a mere 300 billion difficulty all current and proposed mining devices would be operated at a significant loss (assuming $100 exchange rate & $0.10 electrical rate). So at 1 trillion in difficulty even a Cointera rig would be converting $4 in electricity into $1 in Bitcoins. Anyone think that is likely? Significantly higher difficulty is going to require either more efficient hardware, a higher exchange rate, or the average cost of power for the network to decline.
Notes:
[1] Numbers used are from mining hardware comparison where available or based on benchmarks reported by owners/manufacturer. For apples to apples comparison all units are based on wattage at the wall (including all system power and inefficiencies). For devices where total wattage at the wall is unavailable it has been estimated. For unreleased devices ("specs") the efficiency reported by manufacturer is used.
[2] Cointerra did not provide system efficiency "at the wall" however did indicate that chip wattage would be better than 0.55 J/GH, based on that I project the at wall efficiency to be ~0.7 J/GH. Updated to 0.6 J/GH at the chip and 0.75 J/GH at the wall based on official statement. This is based on the following assumptions. Chip power: 0.60 W per GH/s * 2000 GH/s = 1200W @ 0.785 VDC. DC to DC board PSU 90% efficient 1200W/0.9 = 1333W @ 12VDC. Balance of system controller & cooling/fans = 50W @ 12 VDC. Total system DC wattage 1383W @ 12VDC. ATX PSU 90% efficient 13832/0.9 = 1537W @ 120VAC. 1537W / 2000 GH/s = ~0.8 J/GH.
[3] Updated to 1.1 J/GH (https://bitcointalksearch.org/topic/m.3307091). 455W @ 1 VDC running at 502 GH/s. Assume 90% DC to DC conversion and 90% AC power supply efficiency results in estimated 560W at the wall.
[4] The Monarch card is reported by BFL to use 350W. It is not clearly indicated if this is 350W at the chip or the card but I will assume the 350W is the total 12VDC wattage for entire card (to include one or more ASIC processors, host interface, card fan, and DC to DC board PSU. The balance of system is provided by the user so some assumptions are necessary to estimate efficiency at the wall. To be used in a PCIe slot the card requires a "traditional motherboard" with one or more PCIe slot, 100W min is assumed for system and fans. Total DC wattage for single card system is 450W DC. Assuming a 90% efficient ATX PSU 450W DC/0.9 = 500W @ 120V AC. 500W/600GH/s = ~0.8 J/GH. Actual wattage may be significantly higher depending on the components selected by the user.
[5] Given BFL's delivery track record and after the fact changes I don't feel comfortable making any prediction. I would recommend users be very conservative and pad the delivery timeframes provided by BFL.
[6] Due to multiple configurations power efficiency varies by number of modules, batch version, and PSU efficiency. The following efficiencies have been reported 9.4 J/GH (66 GH/s @ 620W), 8.8 J/GH (85 GH/s @ 750W), 8.5 J/GH (82 GH/s @ 700W).
[7] Based on reported efficiency at the wall of less than 350W and nominal performance of greater than 400 GH/s of the BabyJet with single hashing board.
Jump to: