Anybody else watching Mtgox Live?

[Crypt_Current]

S3052 said if it doesn't break 2.8 (which it hasn't yet), it should fall well below 2.

Who's going short from here?

[1713966150]

1713966150

[1713966150]

1713966150

[1713966150]

1713966150

[phorensic]

I'm watching Mtgox live and I see nothing that would get the price up to 2.8.  In fact I can't believe it hasn't gone below $2.0 yet.  Well, it did for a millisecond earlier.

[Crypt_Current]

If that wall gets pulled out, it'll be in free fall

[Jonathan Ryan Owens]

There's around 7200 coins mined a day to consider going forward, owned by many miners worldwide. Very probably those coins will start to enter the market soon, as we are coming toward the end of the month. Because of electric bills, and a lack of bounce back in the BTCUSD price, many will have no choice but to sell and cover mining costs.

Some of those coins will end up short, but with only 1:2.5 leverage and non-USD positions exposed to the underlying value loss when it comes to margin requirements, even at 10% the volatility is going to wipe some of those guys out as well.

Healthy $1 - $1.25 range here we come. GPU mining en mass is going the way of the dinosaur. Does anyone want/need really cheap rendering farms? You're about to get your chance with the coming fire sales.

[Crypt_Current]

There's around 7200 coins mined a day to consider going forward, owned by many miners worldwide. Very probably those coins will start to enter the market soon, as we are coming toward the end of the month. Because of electric bills, and a lack of bounce back in the BTCUSD price, many will have no choice but to sell and cover mining costs.

Some of those coins will end up short, but with only 1:2.5 leverage and non-USD positions exposed to the underlying value loss when it comes to margin requirements, even at 10% the volatility is going to wipe some of those guys out as well.

Healthy $1 - $1.25 range here we come. GPU mining en mass is going the way of the dinosaur. Does anyone want/need really cheap rendering farms? You're about to get your chance with the coming fire sales.

I mine, and I sell only when I think I can buy back more coins cheaper relatively soon.

Also, 5x leverage is still available on bitcoinica, you just have to contact zhou and ask for it.  10x coming soon.

[Crypt_Current]

There's around 7200 coins mined a day to consider going forward, owned by many miners worldwide. Very probably those coins will start to enter the market soon, as we are coming toward the end of the month. Because of electric bills, and a lack of bounce back in the BTCUSD price, many will have no choice but to sell and cover mining costs.

Some of those coins will end up short, but with only 1:2.5 leverage and non-USD positions exposed to the underlying value loss when it comes to margin requirements, even at 10% the volatility is going to wipe some of those guys out as well.

Healthy $1 - $1.25 range here we come. GPU mining en mass is going the way of the dinosaur. Does anyone want/need really cheap rendering farms? You're about to get your chance with the coming fire sales.

Also, as long as I'm in the living situation I'm in, I will NEVER stop mining.   

[phorensic]

Healthy $1 - $1.25 range here we come. GPU mining en mass is going the way of the dinosaur. Does anyone want/need really cheap rendering farms? You're about to get your chance with the coming fire sales.

Actually, mining power has found an equalization point in the last two difficulty changes.  Nothing like the huge drops the 4 changes before it.

[Jonathan Ryan Owens]

Healthy $1 - $1.25 range here we come. GPU mining en mass is going the way of the dinosaur. Does anyone want/need really cheap rendering farms? You're about to get your chance with the coming fire sales.

Actually, mining power has found an equalization point in the last two difficulty changes.  Nothing like the huge drops the 4 changes before it.

Most miners are operating at an unsustainable loss. The appearance of equilibrium from the past 2 drops isn't indicative of anything more than a race to the bottom. The miners that will quit first are those with high electric costs / MH/s, and it will be a rapid drop. Just because the network didn't take a dive on previous drops does not mean that $1 - $1.5 isn't a breaking point. It most certainly is. Overall, the network is underwater on electric costs by about 30% of daily revenue. Over a month, that equals about $200k in loss overall. That is most certainly not going to continue.

[Edward50]

I think most people have been holding off selling when the price dropped to around $3.00. But now that it has been dropping slowly from that price point, I think it will be very soon that people will start to dump their coins.

[proudhon]

I think most people have been holding off selling when the price dropped to around $3.00. But now that it has been dropping slowly from that price point, I think it will be very soon that people will start to dump their coins.

The next drop will be interesting to watch.  I don't think there's any question that we're heading for $1.xx coins.

[Crypt_Current]

I think most people have been holding off selling when the price dropped to around $3.00. But now that it has been dropping slowly from that price point, I think it will be very soon that people will start to dump their coins.

The next drop will be interesting to watch.  I don't think there's any question that we're heading for $1.xx coins.

Hey I thought (MOSTLY FROM YOUR NEW PIC) that you are a bull now

[proudhon]

I think most people have been holding off selling when the price dropped to around $3.00. But now that it has been dropping slowly from that price point, I think it will be very soon that people will start to dump their coins.

The next drop will be interesting to watch.  I don't think there's any question that we're heading for $1.xx coins.

Hey I thought (MOSTLY FROM YOUR NEW PIC) that you are a bull now

Oh, yeah, <cough>, right, <cough>.  Damn, I hope the price goes up.  Er, I mean, the price is going to go up, up, up!  Get ready for it!

[evoorhees]

Because of electric bills, and a lack of bounce back in the BTCUSD price, many will have no choice but to sell and cover mining costs.

I see this statement often, but I question the validity of it. Many people who mine have jobs, and money is fungible. Just because they have an electric bill, and that bill is higher with mining, it doesn't mean they will necessarily sell mined coins to pay that bill. Personally, mining almost doubled my electric bill, but I've never once sold a mined coin. I simply pay the bill with USD and keep the valuable stuff.

[Jonathan Ryan Owens]

Because of electric bills, and a lack of bounce back in the BTCUSD price, many will have no choice but to sell and cover mining costs.

I see this statement often, but I question the validity of it. Many people who mine have jobs, and money is fungible. Just because they have an electric bill, and that bill is higher with mining, it doesn't mean they will necessarily sell mined coins to pay that bill. Personally, mining almost doubled my electric bill, but I've never once sold a mined coin. I simply pay the bill with USD and keep the valuable stuff.

There's a mistake people make when correlating the outside world with their own version of reality. While you may have the resources to pay 2x your regular electric bill, and while you may see bitcoin as valuable in the long term and choose to hoard, there's far more people who have sunk costs and ongoing bills to pay, often on borrowed money, who will be exiting from mining altogether within the next few weeks. It's going to be a good test of Bitcoin to see how well it fares with such a steep drop off in network hashing power.

[bittenbob]

There's around 7200 coins mined a day to consider going forward, owned by many miners worldwide. Very probably those coins will start to enter the market soon, as we are coming toward the end of the month. Because of electric bills, and a lack of bounce back in the BTCUSD price, many will have no choice but to sell and cover mining costs.

Some of those coins will end up short, but with only 1:2.5 leverage and non-USD positions exposed to the underlying value loss when it comes to margin requirements, even at 10% the volatility is going to wipe some of those guys out as well.

Healthy $1 - $1.25 range here we come. GPU mining en mass is going the way of the dinosaur. Does anyone want/need really cheap rendering farms? You're about to get your chance with the coming fire sales.

If GPU mining en masse becomes a thing of the past given current BTC prices then Bitcoin is dead. There is no way anyone would shell out the big bucks for a FPGA board if there is no return. With remaining FPGA boards doing the only mining then a 51% attack becomes a very real problem and then prices will be near zero. Maintaining a minimum $1-2 BTC is essential for the health of the Bitcoin community and centralizing the power in the hands of a few miners would go against everything Bitcoin stands for. This should not be anything anyone should be wishing for.

[trogdorjw73]

Ding ding ding! We have a winner! I suppose we could characterize JRO's statement as him merely wanting BTC to fail, but that doesn't seem to be the case. Really, we should want prices *above* $3, because at that point mining is viable on a scale that would make hacking BTC very difficult. We're nearly there, but only because we have a lot of GPU miners; if they disappear and CPU mining comes back, BTC is dead.

[bittenbob]

Ding ding ding! We have a winner! I suppose we could characterize JRO's statement as him merely wanting BTC to fail, but that doesn't seem to be the case. Really, we should want prices *above* $3, because at that point mining is viable on a scale that would make hacking BTC very difficult. We're nearly there, but only because we have a lot of GPU miners; if they disappear and CPU mining comes back, BTC is dead.

This is why I said in other posts that prices should really be around $3/BTC. Let's hope people start to regain faith in BTC again and start pushing the price more in that direction. You might even wonder if this effort to bring BTC so low is an attempt by some government who dislikes it to bring the currency to halt (ie USA). For those of you cheering for $1 BTC you need to give your heads a shake. Considering the reward per block doesnt drop until 2013 there is a long way to go before the number of bitcoins entering the market starts to drop and theoretically should drive the price up for that reason. If you care about bitcoins you should be encouraging everyone to buy them right now.

On another but possibly related note has there been any discussion on what the effect of the botnet viruses for BTC mining would be and what the hackers purpose with those coins could be?

[RyNinDaCleM]

We're nearly there, but only because we have a lot of GPU miners; if they disappear and CPU mining comes back, BTC is dead.

This will never happen! A CPU uses over half the power (i7 @130w+)and hashes at something around 1/30th the speed. It would be so inefficient it would be pointless.

[ElectricMucus]

The whole mining viability thing for a particular price is bogus. We financed a gpu arms race that's all there is to it.
Security isn't an issue at the current project stage. An attack by any highly funded entity is in the realm of paranoia.  

If any attack were to be attempted it would be a media campaign which is less expensive than a technical attack at this point. As for people investing in power hungry rigs too late: That's the rules of the game, in a gamble someone's gonna pay.

[Cluster2k]

The beautiful design of bitcoin ensures that people dropping out of GPU mining wouldn't kill bitcoin.  The block generation rate remains the same regardless of how many people are mining or what they're mining with.  The only significant danger is miners dropping out en masse in an instant, with a large number of blocks to go before the next difficulty reset.  Unless bitcoin's price drops well under $1 from today's levels I don't really see that happening.  There are enough people mining with 'free' or stolen electricity to keep it going anyway.

[Zotia]

S3052 said if it doesn't break 2.8 (which it hasn't yet), it should fall well below 2.

I predict that if it doesn't go up it will go down.

[proudhon]

Bitcoin Price Forecast. There are 50/50 chances that bitcoin will not make another low below 2 $ and rally from current levels.

Buy!  No.  Sell!  Wait, huh?

[BTCurious]

Bitcoin Price Forecast. There are 50/50 chances that bitcoin will not make another low below 2 $ and rally from current levels.

I predict that if it doesn't go up it will go down.

[BadBear]

It went up 4 cents RALLY!

[DeathAndTaxes]

Overall, the network is underwater on electric costs by about 30% of daily revenue. Over a month, that equals about $200k in loss overall. That is most certainly not going to continue.

You know this how?

~8TH @ 2MH/W = 4000kW load or 96,000 kWh daily.

7200 coins @ $2.25 ea = $16,200 daily.  If that is 30% short then you are saying daily electrical costs are actually $23,100.

Assuming every single miner paid for power directly that is $23,100 / 96,000 = $0.24 per kWh.  

For your "analysis" to be accurate the average global electrical cost would need to be $0.24 per kWh.

1) Average cost in US is $0.10 and US makes up at least a significant portion of the network.
2) The most expensive miners will be first to quit which will drive down average cost.
3) This ignores AC costs but it also ignores the "heat subsidy" in the winter.

So once again how do you "know" that the network is 30% underwater on electrical costs?

[Jonathan Ryan Owens]

Overall, the network is underwater on electric costs by about 30% of daily revenue. Over a month, that equals about $200k in loss overall. That is most certainly not going to continue.

You know this how?

~8TH @ 2MH/W = 4000kW load or 96,000 kWh daily.

7200 coins @ $2.25 ea = $16,200 daily.  If that is 30% short then you are saying daily electrical costs are actually $23,100.

Assuming every single miner paid for power directly that is $23,100 / 96,000 = $0.24 per kWh.  

For your "analysis" to be accurate the average global electrical cost would need to be $0.24 per kWh.

1) Average cost in US is $0.10 and US makes up at least a significant portion of the network.
2) The most expensive miners will be first to quit which will drive down average cost.
3) This ignores AC costs but it also ignores the "heat subsidy" in the winter.

So once again how do you "know" that the network is 30% underwater on electrical costs?

You're brother is Ken, right? You are the other side of the family business - Mining. You're upset about bitcoin devaluation, upset about your Mining operation making virtually nothing. You and others are upset about the rapid decline of interest in bitcoin, and I'd also be upset that cXc launched without as much as a "congratulations" in the forum.

Further, I think you're upset/angry at me specifically for standing up and answering to your friend Jay's slander and libel, and I think you feel utterly retarded for having become involved with (you recommended him to your brother to do business with, remember?) a pathological liar and sociopath.

I think I understand. I'm your nemesis now. It's obvious that you just want to disagree with me. You're overshooting in attacking me though; it's very clear that you are emotionally charged by virtue of these poorly thought out and researched passive aggressive questions. However, your numbers are wrong, and your assumptions are the product of desperation. You really want to believe your own horse shit.

That said, here's my reply.

1) Average cost in the US is $0.15, with most Urban areas having tiered rates much MUCH higher than that. Your reference of $0.10 average was from a mother goose article from 2008, the top result on google. By 2009, for example, average electric costs were already $0.12, and from 2009 onward the additional inflationary pressures on electric rates have been largely absorbed by heavy users. For a single GPU mining, the lower rate probably applies. For anyone running an actual rig or two, they can expect to double their electric bill by stepping up their usage tier.

2) True, and difficulty will drop. The price will drop even further when this happens. Race to the bottom.

3) Wouldn't that be a wash? This is fallacy. In real life though, most people don't keep their 4 rigs in the kitchen. Perhaps the most hardened miiners will strategically move rigs around their house to provide heat for their (very) annoyed families. Maybe.

-added-

4) MANY miners have sunk costs that are borrowed @ interest. Whether it be credit cards, personal loans or otherwise, quite a few fire sales fo inefficient rigs will need to be liquidated in a race to the bottom.

5) Most miners lie about their production vs. costs because none of you want to admit that you're in a geek cult.

6) 7200 coins / day will not get $2.25, but will get some average price lower than that. Expect the increasing pressure to liquidate toward the end of the month to cause the price to gradually decrease like chinese water torture on the charts.

I'm sorry that you're in a rough place, and that you have fallen down the rabbit hole, sir. I had mining rigs for 1 month and sold them at a minor loss, and it was the best decision I ever made. I understand where you're coming from, and I get that you want mining to be profitable again. I know you don't like the bears in this forum, but the reality is that the bears are the only logical people left. The rest of you are out of touch with the current reality.

Take care of yourself.

-Jonathan

[BadBear]

You're brother is Ken, right? You are the other side of the family business - Mining. You're upset about bitcoin devaluation, upset about your Mining operation making virtually nothing. You and others are upset about the rapid decline of interest in bitcoin, and I'd also be upset that cXc launched without as much as a "congratulations" in the forum.

Further, I think you're upset/angry at me specifically for standing up and answering to your friend Jay's slander and libel, and I think you feel utterly retarded for having become involved with (you recommended him to your brother to do business with, remember?) a pathological liar and sociopath.

I think I understand. I'm your nemesis now. It's obvious that you just want to disagree with me. You're overshooting in attacking me though; it's very clear that you are emotionally charged by virtue of these poorly thought out and researched passive aggressive questions. However, your numbers are wrong, and your assumptions are the product of desperation. You really want to believe your own horse shit.

That said, here's my reply.

1) Average cost in the US is $0.15, with most Urban areas having tiered rates much MUCH higher than that. Your reference of $0.10 average was from a mother goose article from 2008, the top result on google. By 2009, for example, average electric costs were already $0.12, and from 2009 onward the additional inflationary pressures on electric rates have been largely absorbed by heavy users. For a single GPU mining, the lower rate probably applies. For anyone running an actual rig or two, they can expect to double their electric bill by stepping up their usage tier.

2) True, and difficulty will drop. The price will drop even further when this happens. Race to the bottom.

3) Wouldn't that be a wash? This is fallacy. In real life though, most people don't keep their 4 rigs in the kitchen. Perhaps the most hardened miiners will strategically move rigs around their house to provide heat for their (very) annoyed families. Maybe.

-added-

4) MANY miners have sunk costs that are borrowed @ interest. Whether it be credit cards, personal loans or otherwise, quite a few fire sales fo inefficient rigs will need to be liquidated in a race to the bottom.

5) Most miners lie about their production vs. costs because none of you want to admit that you're in a geek cult.

6) 7200 coins / day will not get $2.25, but will get some average price lower than that. Expect the increasing pressure to liquidate toward the end of the month to cause the price to gradually decrease like chinese water torture on the charts.

I'm sorry that you're in a rough place, and that you have fallen down the rabbit hole, sir. I had mining rigs for 1 month and sold them at a minor loss, and it was the best decision I ever made. I understand where you're coming from, and I get that you want mining to be profitable again. I know you don't like the bears in this forum, but the reality is that the bears are the only logical people left. The rest of you are out of touch with the current reality.

Take care of yourself.

-Jonathan

[DeathAndTaxes]

nonesense snipped.

I have no idea who Ken is.  I have no idea who Jay is either.


You made a very simple claim.  

Overall, the network is underwater on electric costs by about 30% of daily revenue.

First that statement is easily verified as false.  Second nothing in your ranting response addresses the metrics.

For your claim to be correct at only 2MH/W it would require an average electrical rate of $0.24.

The $0.10 I was going on was based on memory.  Looks like it is now $0.13 now.  
Still even if it was $0.15 or $0.18 or $0.20 you would still be wrong.  
http://www.bls.gov/ro5/aepchi.htm




Also I used 2.0 MH/W to be ultra conservative (and give you the largest benefit of doubt).  At a more realistic 2.2 MH/W it is closer to average global electrical rate to be $0.26 for your claim to be accurate.

To address your non-points.
1) Doubling the electric bill isn't relevent.  What matters is the cost of electricity which is nowhere near $0.25 per kWh.  The price in urban vs rural isn't relevent either as the average price is what matters.

2) Which doesn't change the fact that AVERAGE ELECTRICAL COST WOULD GO DOWN!  If difficulty and price both fall it doesn't change the fact that the marginal miners would be the one's most likely to quit.  In other words the average rate of the quitters would be higher than the network average so as hashing power falls the network average electrical rate would also decline.  The opposite is also true as hashing power increases as more marginal miners believe they can "make a profit".

3) I was indicating it is a wash but still it has nothing to do w/ your claim of 30% below electrical cost.

4) Which has nothing to do w/ your claim of 30% below electrical cost

5) Which has nothing to do w/ your claim of 30% below electrical cost

6) Of course it will.  Miners sell coins everyday.  Electricity can't be paid in BTC so coins are sold to pay electrical costs.  

I'm sorry that you're in a rough place, and that you have fallen down the rabbit hole, sir. I had mining rigs for 1 month and sold them at a minor loss, and it was the best decision I ever made. I understand where you're coming from, and I get that you want mining to be profitable again. I know you don't like the bears in this forum, but the reality is that the bears are the only logical people left. The rest of you are out of touch with the current reality.

I never said the price is going up I think price is very likely to decline.  As a miner I don't really care about price at all, only price relative to difficulty.  This isn't a bull vs bear debate.  It is a you making false claim, getting busted, and then throwing a temper tantrum.

I don't need mining to be profitable again.  My rigs are paid for and my electrical cost (at current difficulty) is $0.88 per BTC (@ current difficulty).  The best thing that could happen for me would be the marginal and unprofitable miners to be washed out.  If price were to fall to $0.88 per BTC difficulty would also fall making my costs fall.

[DeathAndTaxes]

Ding ding ding! We have a winner! I suppose we could characterize JRO's statement as him merely wanting BTC to fail, but that doesn't seem to be the case. Really, we should want prices *above* $3, because at that point mining is viable on a scale that would make hacking BTC very difficult. We're nearly there, but only because we have a lot of GPU miners; if they disappear and CPU mining comes back, BTC is dead.

Why would GPU mining disapear at any price point.  GPUs are more efficient than CPUs.  Period.  They produce more MH/$ and more MH/W.

If 90% of miner's quit then difficulty will fall 90% but even then it wouldn't make sense to use CPU.  The trend will always be toward more efficient hardware.

Low efficiency CPU mining -> OpenCL CPU Mining -> GPU Mining -> FPGA Mining -> Structured ASICS mining -> Full Custom ASIC mining.

$10,000 per BTC or $0.0001 per BTC it always makes sense to use the most efficient gear.

[Jonathan Ryan Owens]

Ding ding ding! We have a winner! I suppose we could characterize JRO's statement as him merely wanting BTC to fail, but that doesn't seem to be the case. Really, we should want prices *above* $3, because at that point mining is viable on a scale that would make hacking BTC very difficult. We're nearly there, but only because we have a lot of GPU miners; if they disappear and CPU mining comes back, BTC is dead.

Why would GPU mining disapear at any price point.  GPUs are more efficient than CPUs.  Period.  They produce more MH/$ and more MH/W.

If 90% of miner's quit then difficulty will fall 90% but even then it wouldn't make sense to use CPU.  The trend will always be toward more efficient hardware.

Low efficiency CPU mining -> OpenCL CPU Mining -> GPU Mining -> FPGA Mining -> Structured ASICS mining -> Full Custom ASIC mining.

$10,000 per BTC or $0.0001 per BTC it always makes sense to use the most efficient gear.

I wasn't throwing a temper tantrum. I do agree with you about efficient gear, though. I also don't think cryptocurrencies are going to die. For instance, if BFL is for real, then they have already achieved the 5th step you referenced, and will be able to deliver performance that will curb the outflow of fiat to the electric companies. I should have stated further that GPU mining will disappear because of these upcoming options. 1/100th the electric cost is a good thing for Bitcoin.

I'm sorry that I assumed you were involved / related to Ken, or had any knowledge of the situation referenced. Your response was measured and calm, and I respect you for that.

My real concern is that there is a network that costs $20k per day in electricity to operate, for a network that currently produces very little value. I'm not suggesting that Bitcoin or its variants won't have a place in the world economy in a few years, but I am suggesting that the network is oversubscribed, and that it's a race to the bottom, where desperation and greed have won out over logic and reason.

As to your question about being 30% underwater, here's my assumptions with 0 cost for hardware and $0.135 cents per Kw/h. It's almost exactly 30%, by the way (once I did the maths).

1000 MH/s @ 750W
Electric Rate: $0.135
Price: $2.26
Time Period: 1 Month
Difficulty: 1192497

Results:

Electric cost for 1 month time frame: $74
Revenue per time frame: $56
P/L: -$18 USD [edit: bad maths]
Percentage Loss: 32%

So, in order for mining to be at break even edit: at current prices, difficulty needs to be at around 800,000. At a difficulty of 800,000, and assuming that there's correlation between the network and price (an oft debated question), the price at 800,000 will be around $1, which means that miners will be MORE underwater than they are now. Lowering difficulty will not result in higher incomes, but rather will cause even steeper declines, which again, will result in greater attrition. It's a race to 0.

[trogdorjw73]

So I'm curious: what sort of hardware (and clocks) are people running to get anywhere near 2MH/s per Watt? Because I've got a system with overclocked 5870 and 5950 on an 850W 80 Plus Gold PSU, and the whole system is drawing 570W and producing 760 Mhash/s. That's a pretty massive difference between a "conservative" 2Mh/W and my result of 1.35Mh/W. I'd need to cut my system's power use to just 380W to reach your claimed 2Mh/W, and I don't see that happening with the current hardware. FPGA hashing could probably get there, but GPU mining? I'm not sure how anyone would get that low. Got a good link for suggested optimizations?

[Jonathan Ryan Owens]

So I'm curious: what sort of hardware (and clocks) are people running to get anywhere near 2MH/s per Watt? Because I've got a system with overclocked 5870 and 5950 on an 850W 80 Plus Gold PSU, and the whole system is drawing 570W and producing 760 Mhash/s. That's a pretty massive difference between a "conservative" 2Mh/W and my result of 1.35Mh/W. I'd need to cut my system's power use to just 380W to reach your claimed 2Mh/W, and I don't see that happening with the current hardware. FPGA hashing could probably get there, but GPU mining? I'm not sure how anyone would get that low. Got a good link for suggested optimizations?

Hint: They're lying.

[Crypt_Current]

assuming that there's correlation between the network and price (an oft debated question)

That's the weak point of your argument.  Many miners mine off "free" electricity for the sake of hoarding over the long run, with absolutely no correlation to costs in USD.  I'm sure some even get their hardware at zero effective cost, as gifts for example.

[Crypt_Current]

So I'm curious: what sort of hardware (and clocks) are people running to get anywhere near 2MH/s per Watt? Because I've got a system with overclocked 5870 and 5950 on an 850W 80 Plus Gold PSU, and the whole system is drawing 570W and producing 760 Mhash/s. That's a pretty massive difference between a "conservative" 2Mh/W and my result of 1.35Mh/W. I'd need to cut my system's power use to just 380W to reach your claimed 2Mh/W, and I don't see that happening with the current hardware. FPGA hashing could probably get there, but GPU mining? I'm not sure how anyone would get that low. Got a good link for suggested optimizations?

might happen with the next series of radeons

[DeathAndTaxes]

So I'm curious: what sort of hardware (and clocks) are people running to get anywhere near 2MH/s per Watt? Because I've got a system with overclocked 5870 and 5950 on an 850W 80 Plus Gold PSU, and the whole system is drawing 570W and producing 760 Mhash/s. That's a pretty massive difference between a "conservative" 2Mh/W and my result of 1.35Mh/W. I'd need to cut my system's power use to just 380W to reach your claimed 2Mh/W, and I don't see that happening with the current hardware. FPGA hashing could probably get there, but GPU mining? I'm not sure how anyone would get that low. Got a good link for suggested optimizations?

Rig needs to be optimized not a bunch of crap thrown together.

I have 5 identical rigs:
MSI AM3 MB
Sepron 145
2x120mm fans
3x5970 @ 850/160 and stock voltage.
Linux coin on usb stick.
EVERYTHING turned off in bios (USB 3.0, SATA, onboard audio, IDE controller, etc)
Underclock sempron to 2.0Ghz (likely could go lower).
Don't overvolt GPU, be sure to undervolt memory.
80-Plus GOLD powersupply.

I get 2.25GH per rig @ 870W from the wall. ~11GH total on 4.4KW. If you are getting sub 2 MH/W you simply can't compete unless you have "free" power.  Even if you "could" compete it is throwing away thousands of dollars (in any significant size rig/farm) for no extra benefit.  I am considering running a 240V line in the garage to squeeze another 3% efficiency out of the PSU.  If the prices drop below $2 I might consider lowering the overclock of 5970s to undervolt them and gain some efficiency (change in power is change in voltage squared).

Bitcoins are a commodity business.  Price is set by the lowest cost producer.    Price will eventually stabilize at maybe 10% ROI over the efficient producer's cost.  

Honestly I have no idea how you are pulling 570W from that rig.  Even a pair of 5870s should be a lot less than that.  BTW I don't think they make a 5950.

100% cpu bug on quadcore Intel running 24/7?
Not underclocking memory?

[Jonathan Ryan Owens]

assuming that there's correlation between the network and price (an oft debated question)

That's the weak point of your argument.  Many miners mine off "free" electricity for the sake of hoarding over the long run, with absolutely no correlation to costs in USD.  I'm sure some even get their hardware at zero effective cost, as gifts for example.

Weak point? Seriously? I think that you and far too many others are suffering from an obsessive belligerence towards reality. Do you honestly think Visa/Mastercard are successful because they can steal electricity? Do you suppose that Bitcoin acting as a proxy to steal from mary (mom and dad / work) to pay peter is a solid foundation which will give Bitcoin the 'edge' it needs? Lol..

Lots of two dimensional thinking from the lot of you, where ALL or NONE seems to be your response to everything. Yes, there are a % of miners who don't pay for their electricity, but those costs are passed on to someone. The vast majority, however, will either get caught by mom and dad, or will liquidate their mining equipment. The price isn't going back up, sir. It is on a continued decline. GPU mining is not here to stay, and that's a good thing. What a waste of perfectly good video cards, and electricity generated mostly from coal.

[ultima]

That's the weak point of your argument.  Many miners mine off "free" electricity for the sake of hoarding over the long run, with absolutely no correlation to costs in USD.  I'm sure some even get their hardware at zero effective cost, as gifts for example.

I bought my mining rig and I pay my electricity bills, but I will not stop mining. One reason is contribution to the network, the other is long term view. I don't care if my daily mining production barely covers my current electricity costs and I don't care if at some point price falls below my mining cost. I have enough fiat paper to pay a few more € for additional electricity.

For someone who borrowed money, bought mining rig and hoped to get rich quick... well, they have a problem.

[DeathAndTaxes]

As to your question about being 30% underwater, here's my assumptions with 0 cost for hardware and $0.135 cents per Kw/h. It's almost exactly 30%, by the way (once I did the maths).

1000 MH/s @ 750W
Electric Rate: $0.135
Price: $2.26
Time Period: 1 Month
Difficulty: 1192497

Well at least I now understand where you are coming from.  I 100% agree that if every miner is that inefficient the network is operating at a loss.  Hopefully they shutoff their rigs.  1000MH / 750W = 1.33MH/W.  Can't compete.  Will never be able to compete.  People w/ rigs or prices like that should just shut their rigs off and pay someone who can get better efficiency and power prices to mine for them.  I agree the network currently does not need 8TH of security not to protect the tiny amount of true transaction volume.  People who refuse to admit they can't compete or think they are "helping the network" aren't helping anyone.  They are holding back market forces which will eventually reach a price/difficulty ratio that allows for a modest profit margin for those who have both the electrical rates and efficiency to compete.

BTW same numbers but at 2.6 MH/W and $0.08 per kWh = $28 electrical cost or $48 profit not $18 loss.  At 11GH that is ~$500 per month.

Still I disagree with falling difficulty "helping".  it wouldn't matter if difficult was 800K if you are that inefficient.  If difficulty relative to price (the only metric a miner should care about) fell that low more efficient miners would add capacity and drive inefficient miners into the red.  Bitcoins are a commodity business.  You can't compete with poor efficiency and higher input costs in a commodity business.  It is true at a difficulty of 100K or 500M. 

The speculative bubble made it impossible for anyone w/ any rigged up junk to avoid falling into piles of cash.  I mean you couldn't be bad enough to not make 100%, 200%, 500% annualized ROI when BTC was $30 but that wasn't sustainable.  

Lets take you example miner 1000MH @ 750W & $0.135 per kWh.  (Lets say rig was poorly designed and costs $1.5 per MH)
At $30 BTC and 2 million difficulty that is
Gross: $452.62
Electrical cost: - $74
Ammortized hardware (3 years):  -$43
Net: $335.62 per month or $4027 annually on a $1500 investment.
ROI%:

No economy/industry allows the worst producers 200%+ ROI.  If BTC had stayed at $30 difficulty would have eventually been 20M making those same miners operated at the same loss they would at current prices & difficulty.

[Jonathan Ryan Owens]

As to your question about being 30% underwater, here's my assumptions with 0 cost for hardware and $0.135 cents per Kw/h. It's almost exactly 30%, by the way (once I did the maths).

1000 MH/s @ 750W
Electric Rate: $0.135
Price: $2.26
Time Period: 1 Month
Difficulty: 1192497

Well at least I now understand where you are coming from.  I 100% agree that if every miner is that inefficient the network is operating at a loss.  Hopefully they shutoff their rigs.  1000MH / 750W = 1.33MH/W.  Can't compete.  Will never be able to compete.  People w/ rigs or prices like that should just shut their rigs off and pay someone who can get better efficiency and power prices to mine for them.

Same numbers but at 2.6 MH/W and $0.08 per kWh = $28 electrical cost or $48 profit not $18 loss. 

It wouldn't matter if difficult was 800K if you are that efficient.  If difficulty relative to price fell that low more efficient miners would add capacity and drive inefficient miners into the red.  Bitcoins are a commodity business.  You can't compete with poor efficiency and higher input costs in a commodity business.

The speculative bubble made it impossible for anyone w/ a bunch of rigged up trash to not make money.  I mean you couldn't be bad enough to not make 100%, 200%, 500% annualized ROI when BTC was $30 but that wasn't sustainable.  No economy/industry allows the worst producers to make multi hundreds of % ROI.  If BTC had stayed at $30 difficulty would have eventually been 20M.

If you're getting 2.25 MH/W, I commend you. I also need to point out that the vast majority of miners are not so lucky, and I'd go so far as to assume that 30-40% of current mining capacity is running not only in the red, but at the numbers that I mentioned above. This is an inescapable reality that is soon going to reduce the total hashing power by at least 1/3, because while there are those like ultima who have extra fiat and a sense of responsibility to help Bitcoin survive, there's also many who got into mining because they thought they'd finally be able to afford a private island.

-Jonathan

[DeathAndTaxes]

If you're getting 2.25 MH/W, I commend you. I also need to point out that the vast majority of miners are not so lucky, and I'd go so far as to assume that 30-40% of current mining capacity is running not only in the red, but at the numbers that I mentioned above. This is an inescapable reality that is soon going to reduce the total hashing power by at least 1/3, because while there are those like ultima who have extra fiat and a sense of responsibility to help Bitcoin survive, there's also many who got into mining because they thought they'd finally be able to afford a private island.

Well luck has nothing to do with it.  

Simple economics and understanding of commodity based businesses.  If you want to compete in a commodity based business you need to have the lowest cost.  Period.

I agree about hashing power falling.  My guesstimate is that hashing power is closer to 5TH by next summer.   If there are that many miners operating that far in the red well ... they aren't helping themselves, they aren't helping bitcoin, and they aren't even being good speculators.

If your cost to produce a coin is $3 and you can buy one for $2.50 even if you honestly believe that someday BTC will be $10,000 why not just buy the coins?  If BTC DOES got to $10,000 well you ended up buying 20% more for the same cost.  20% more millions.  Right?

Market forces always win in the end.  Those propping up hashing power at a loss aren't helping Bitcoin.   If Bitcoin is sucessful prices will rise and when they rise it will be possible to more network capacity to exist.  For the tiny economy that Bitcoin has 1TH is more than sufficient.  8TH is simply burning 8x the electricity for no gained utility.

[proudhon]

Bitcoin Price Forecast. There are 50/50 chances that bitcoin will not make another low below 2 $ and rally from current levels.

Oh my god, he was right!

[Crypt_Current]

Bitcoin Price Forecast. There are 50/50 chances that bitcoin will not make another low below 2 $ and rally from current levels.

Oh my god, he was right!

Come on now, that's Twitter -- you have to have filler for Twitter.  I don't even follow tweets for this reason.
Besides, this is old news... newest forecast is 80% "down hard"