Huge ~40% Drop in Network Hashing Power (5/27/2011)

[xenon481]

The total network hashing power started 5/27/2011 UTC at ~5THash/s and swung to close the day at just over ~3THash/s. With a low around ~2.9THash/s.

That is a nearly 40% drop, day over day, in the first full day after the latest 78% difficulty increase.

Edit: This is based on the graphs from http://bitcoin.sipa.be/ which do not have the same calculation flaw right after a difficulty change that BitcoinCharts does.


Uploaded at ImageShack.

[frozen]

btcguild was down, it seems to be back up now.

[xenon481]

btcguild was down, it seems to be back up now.

BTCGuild alone doesn't account for a ~2THash swing, though.

[Dobrodav]

Bitcoinchart crude assumption on total hashrate,  and mainly - psychological  factor of that assumptions on fresh miners did that.
Many is quit from mining.

[frozen]

btcguild was down, it seems to be back up now.

BTCGuild alone doesn't account for a ~2THash swing, though.

No, but also remember that hashing power is calculated based on the current difficulty and the last few blocks. So when there is a dramatic increase in difficulty, this throws off the calculation, which can explain why the hash power shot up above 5Thash and then back down to 3.4Thash. btcguild was doing almost 300Ghash, which would put the network at 3.1Thash, which appears to be where it dropped to.

[xenon481]

Bitcoinchart crude assumption on difficulty, total hashrate, [...]

No, but also remember that hashing power is calculated based on the current difficulty and the last few blocks. So when there is a dramatic increase in difficulty, this throws off the calculation, which can explain why the hash power shot up above 5Thash and then back down to 3.4Thash. btcguild was doing almost 300Ghash, which would put the network at 3.1Thash, which appears to be where it dropped to.

The sipa charts are not based on BitcoinChart's network hashrate calculations which are flawed around the time after a difficulty change.

http://bitcoin.sipa.be/

Edit: The sipa charts do not suffer from the same flawed calculation methodology that bitcoincharts does.

[Dobrodav]

So? dit that sipa charts show any drop in network hashrate ? Or they just to much flawed around difficulty raises ?

[xenon481]

So? dit that sipa charts show any drop in network hashrate ? Or they just to much flawed around difficulty raises ?

The sipa graph is the graph that I posted in the OP. It shows a 40% intraday swing. If you look at some of the other graphs on the sipa site, you will see that it those graphs do not have the same difficulty change miscalculation as bitcoincharts.

[EPiSKiNG]

The total network hashing power started 5/27/2011 UTC at ~5THash/s and swung to close the day at just over ~3THash/s. With a low around ~2.9THash/s.

That is a nearly 40% drop, day over day, in the first full day after the latest 78% difficulty increase.

YAY!!!!!!!

-EP

[jasonk]

The total network hashing power started 5/27/2011 UTC at ~5THash/s and swung to close the day at just over ~3THash/s. With a low around ~2.9THash/s.

That is a nearly 40% drop, day over day, in the first full day after the latest 78% difficulty increase.

Edit: This is based on the graphs from http://bitcoin.sipa.be/ which do not have the same calculation flaw right after a difficulty change that BitcoinCharts does.


Uploaded at ImageShack.

Is this a joke?  This happens everytime the difficulty increases.... the accuracy of the hash rate calculation is not correct because of the 24hr sample rate.  I would assume someone with a few hundred posts on the forum would know better....

[Sweft]

The total network hashing power started 5/27/2011 UTC at ~5THash/s and swung to close the day at just over ~3THash/s. With a low around ~2.9THash/s.

That is a nearly 40% drop, day over day, in the first full day after the latest 78% difficulty increase.

Edit: This is based on the graphs from http://bitcoin.sipa.be/ which do not have the same calculation flaw right after a difficulty change that BitcoinCharts does.


Uploaded at ImageShack.

Is this a joke?  This happens everytime the difficulty increases.... the accuracy of the hash rate calculation is not correct because of the 24hr sample rate.  I would assume someone with a few hundred posts on the forum would know better....

Actually the green lone is an 8 hour moving average.

[xenon481]

Is this a joke?  This happens everytime the difficulty increases.... the accuracy of the hash rate calculation is not correct because of the 24hr sample rate.  I would assume someone with a few hundred posts on the forum would know better....

Please read the thread.

Edit: This is based on the graphs from http://bitcoin.sipa.be/ which do not have the same calculation flaw right after a difficulty change that BitcoinCharts does.

The sipa charts are not based on BitcoinChart's network hashrate calculations which are flawed around the time after a difficulty change.

http://bitcoin.sipa.be/

Edit: The sipa charts do not suffer from the same flawed calculation methodology that bitcoincharts does.

The sipa graph is the graph that I posted in the OP. It shows a 40% intraday swing. If you look at some of the other graphs on the sipa site, you will see that it those graphs do not have the same difficulty change miscalculation as bitcoincharts.

[SgtSpike]

This is... GREAT news for those of us still mining!  Hopefully this is a sign of difficulty beginning to level off... I doubt it will stay that low, but I can hope, can't I? 

[timmmay]

Indeed it is good for those of us in it for the long haul!  I was hoping this would happen

[jkminkov]

probably some miner finally managed to melt his substation/transformer and he is alone in the dark

[chungenhung]

deepbit.net is down! that is why.

[SgtSpike]

deepbit.net is down! that is why.

It hasn't been for me... been getting normal rates.

[jasonk]

Please read the thread.

I did.  I still don't believe that chart is accurate.  Something is wrong.  Why else would the difficulty level drop a large amount EXACTLY when difficulty level increased?

[nathanrees19]

Why else would the difficulty level drop a large amount EXACTLY when difficulty level increased?

Assuming you mean why the hashing power would drop, there are dozens of possible explanations.

[Chucksta]

LOL, don't get your hopes up, this happens every time the difficulty increases. Just wait till the next difficulty increase, another 60% or more, I bet.

[SgtSpike]

Please read the thread.

I did.  I still don't believe that chart is accurate.  Something is wrong.  Why else would the difficulty level drop a large amount EXACTLY when difficulty level increased?

Probably because people are saying "screw it", and turning off their miners.

It doesn't make sense if you look at the numbers, as mining is still very profitable, but threads here have already shown that some people are no longer mining due to the difficulty.

[error]

I suspect a lot of the casual miners who came in with the last big round of press hits are getting out. Those who were CPU mining for a month or two and made less than they spent on electricity, or the NVIDIA miners who barely broke even, or even lost. At this difficulty such mining isn't really worth it.

And, summer hit with a vengeance here in the US on the day before the difficulty rose. I suspect some of the miners just can't stand the heat anymore.

[molecular]

Please read the thread.

I did.  I still don't believe that chart is accurate.  Something is wrong.  Why else would the difficulty level drop a large amount EXACTLY when difficulty level increased?

The difficulty increase was around 5/26 6pm, the hashrate was still climbing until after midnight that day. That's HARDLY exactly at the same time.

I assume this is just a lot of smaller miners making recalculations after the strong difficulty increase and deciding mining is not profitable for them any more and they don't want to bet on rising exchange rate, so they just quit or head back to folding@home or seti or whatever.

[Chucksta]

Please read the thread.

I did.  I still don't believe that chart is accurate.  Something is wrong.  Why else would the difficulty level drop a large amount EXACTLY when difficulty level increased?

Probably because people are saying "screw it", and turning off their miners.

It doesn't make sense if you look at the numbers, as mining is still very profitable, but threads here have already shown that some people are no longer mining due to the difficulty.

Exactly, mining does still bring in money, more than enough to cover costs. People are not leaving en masse. I'd be willing to bet (I don't gamble... phobia!) that more are coming in than are leaving.

Reading this thread it strikes me that a lot of people are overly optimistic, thus blinded to any explanations that refute what is shown in the chart.

[SgtSpike]

I suspect a lot of the casual miners who came in with the last big round of press hits are getting out. Those who were CPU mining for a month or two and made less than they spent on electricity, or the NVIDIA miners who barely broke even, or even lost. At this difficulty such mining isn't really worth it.

And, summer hit with a vengeance here in the US on the day before the difficulty rose. I suspect some of the miners just can't stand the heat anymore.

Mmmm, good point about the heat.  I have a friend in CA who is jealous of my mining, and he has a card that would do 100MH/s at least, but he won't mine because he said his room is already too hot.

Those of us who are adults can just go out and buy something to cool the room with, but some younger miners might not have that option available, and would just shut down the miners instead.

[error]

It got up to 86 here in New Hampshire the past few days. It's HOT.

[molecular]

is there any chance to see a difficulty drop ?

It's happened before, last time when "misteryminer" left in March. see http://bitcoin.sipa.be.

I doubt it will happen this time, though

[Sukrim]

It might be more psychologically relevant:

"Oh, a lot of ppl. are dropping out, I should drop out too!" 

[ChaosFox]

Seems like it's going up anyway.  But, consider this: There's someone, or a group of people with a considerable part of the network's computing power, in the order of the Thash/s.  They only allocate some of that power right before difficulty increases and turn it off right after difficulty increases, taking advantage of the screwiness for the first blocks, in hopes others miners will leave and the difficulty increase forecast actually gets skewed, and as the change approaches and some people leave, they gradually put their power in, ensuring a steady profit. Maybe I'm just nuts, and yes, this is one pretty tinfoil hat.

[k]

Seems like it's going up anyway.  But, consider this: There's someone, or a group of people with a considerable part of the network's computing power, in the order of the Thash/s.  They only allocate some of that power right before difficulty increases and turn it off right after difficulty increases, taking advantage of the screwiness for the first blocks, in hopes others miners will leave and the difficulty increase forecast actually gets skewed, and as the change approaches and some people leave, they gradually put their power in, ensuring a steady profit. Maybe I'm just nuts, and yes, this is one pretty tinfoil hat.

i'm pretty sure it doesn't work like this. how can these mystery miners take advantage of the "screwiness for the first blocks" after the difficulty change as you call it. It's only a screwiness in the calculated hash rate stats that happens after the difficulty change. The difficulty is set for a 2016 block period, it doesn't change. So for a certain hash rate you should get the same return on average for a particular difficulty regardless if it's just right after that difficulty has been set or if it's just before the difficulty is going to be readjusted

[commlinx]

I got scammed by this too, recently I bought one of the cheap ASIC miners on ebay that worked great for a day or two then I got an e-mail it would stop working at block 127225 unless I paid a further $1000 per core for an unlock code. I refused the extortion and guess many others were the same. At 45 Ghash/s for 16 cores it might have been worthwhile but I couldn't be sure I wouldn't be scammed again at some future block count

[Dhomochevsky]

What? Care to elaborate on that? More details about the scam and the scammers please.

[Chucksta]

Seems like it's going up anyway.  But, consider this: There's someone, or a group of people with a considerable part of the network's computing power, in the order of the Thash/s.  They only allocate some of that power right before difficulty increases and turn it off right after difficulty increases, taking advantage of the screwiness for the first blocks, in hopes others miners will leave and the difficulty increase forecast actually gets skewed, and as the change approaches and some people leave, they gradually put their power in, ensuring a steady profit. Maybe I'm just nuts, and yes, this is one pretty tinfoil hat.

i'm pretty sure it doesn't work like this. how can these mystery miners take advantage of the "screwiness for the first blocks" after the difficulty change as you call it. It's only a screwiness in the calculated hash rate stats that happens after the difficulty change. The difficulty is set for a 2016 block period, it doesn't change. So for a certain hash rate you should get the same return on average for a particular difficulty regardless if it's just right after that difficulty has been set or if it's just before the difficulty is going to be readjusted

LOL, the guy is talking out of his butt... on some funny stuff and living in pixie land, LOL

[Jaime Frontero]

I got scammed by this too, recently I bought one of the cheap ASIC miners on ebay that worked great for a day or two then I got an e-mail it would stop working at block 127225 unless I paid a further $1000 per core for an unlock code. I refused the extortion and guess many others were the same. At 45 Ghash/s for 16 cores it might have been worthwhile but I couldn't be sure I wouldn't be scammed again at some future block count

huh?

yes, please elaborate...

[Nythain]

And, summer hit with a vengeance here in the US on the day before the difficulty rose. I suspect some of the miners just can't stand the heat anymore.

Actual temperatures aside... Tornado season hit with a vengeance throughout a pretty hefty chunk of the US... I'm curious how many miners have been without power or even worse, a home, in the last week or two.

[TTBit]

Implied network power for last X blocks, using block explorer:

(2 ^ 256 / (2^224/Difficulty)) / (time in seconds between blocks) * (number of blocks)

Example:
Implied network power between blocks 127259 and 127295

(2 ^ 256 / (2 ^ 224 / 434877.045527)) / (20423 seconds) * (36 blocks) = 3,292,375,105,494.02 hashes per second

(actually, this seems low)

[w128]

Don't worry guys. Im away for the holiday weekend. I had some concern about burning my house down so I shut the miners down. Be back in full swing on Monday midday with new capacity arriving Wednesday.

[allinvain]

What? Wait a minute. Are you saying that you control that hashing power? If so what are you running 1000 5970s?

[SgtSpike]

What? Wait a minute. Are you saying that you control that hashing power? If so what are you running 1000 5970s?

He doesn't, lol.  People are just saying stuff to get attention, but they're all a bunch of liars.

[w128]

What? Wait a minute. Are you saying that you control that hashing power? If so what are you running 1000 5970s?

He doesn't, lol.  People are just saying stuff to get attention, but they're all a bunch of liars.

The word is sarcasm. No need to be an ass.

[SgtSpike]

What? Wait a minute. Are you saying that you control that hashing power? If so what are you running 1000 5970s?

He doesn't, lol.  People are just saying stuff to get attention, but they're all a bunch of liars.

The word is sarcasm. No need to be an ass.

How am I being an ass?

[Chucksta]

What? Wait a minute. Are you saying that you control that hashing power? If so what are you running 1000 5970s?

He doesn't, lol.  People are just saying stuff to get attention, but they're all a bunch of liars.

The word is sarcasm. No need to be an ass.

How am I being an ass?

I think this is a case of people not seeing when others are joking... not easy to tell when you are dealing with text.

Anyway, never mind being an ass, I reckon you're all arses, LOL

/slap me

[Sukrim]

Looks like more tahn just a short down-spike atm! 

Currently, we're below 3,5 TH/s.

Looks like we scared the noobs away! 

[SgtSpike]

AND, this is the first time I've seen the block average at bitcoincharts.com under 6/hr!

[Sukrim]

Time to play http://www.youtube.com/watch?v=f62pW0J2JsA in the background...

[rezin777]

AND, this is the first time I've seen the block average at bitcoincharts.com under 6/hr!

CRAY was testing their new supercomputer on the Bitcoin network and realized they were losing money because it has nVidia GPUs?  

[zale]

Lets hope the BTC value doesn't crash too much. slush's pools Hashrate dropped from 580Ghash/s to 520 (still 95Mhash per worker..wow) and MtGox vrom 8.8 to 8.1$ because everyone of those 50Mhash/s guys seems to sell their stuff.

[Isepick]

See, here in the US it is a 3 day holiday weekend. The Federal employees tasked with running the Fed's FPGA arrays simply turned them off and went home for the weekend. Don't worry, they'll come back to work on Tuesday, turn them back on, and catapult us to a 700K difficulty level.

[benjamindees]

See, here in the US it is a 3 day holiday weekend. The Federal employees tasked with running the Fed's FPGA arrays simply turned them off and went home for the weekend. Don't worry, they'll come back to work on Tuesday, turn them back on, and catapult us to a 700K difficulty level.

That can't possibly be right.  Federal employees are nowhere near competent enough to operate an FPGA.  They would have outsourced it to contractors who are probably working overtime eating Doritos at $300/hr.

[bcpokey]

On the upside I don't have to burn all my mining hardware for kindling to keep myself and my children warm at night now that I spent all our money on it and difficulty made me a poor man...

On the downside (a little more seriously) transactions are taking a lot longer now to process than before. An impatient man who is too cheap to add a transaction fee is saddened by this.

[mathx]

Mmmm, good point about the heat.  I have a friend in CA who is jealous of my mining, and he has a card that would do 100MH/s at least, but he won't mine because he said his room is already too hot.

Those of us who are adults can just go out and buy something to cool the room with, but some younger miners might not have that option available, and would just shut down the miners instead.

Adults like to lose money?

The only thing that determines profitability isnt how absolutely fast your rig is, but how fast your rig is compared to the total hash rate. Bitcoins increase in value relatively in step with difficulty of course, so you get more for more difficult coins. What matters is your chance of being the one to find it/being part of a pool that does.

The more people that drop off, the better it is for all miners, regardless of difficulty. In fact, mtgox might even show value vs other currencies increasing for BTC because its so hard to generate more - the money supply just slowed down, value goes up. And on top of that your chances just increased because people left - best time to mine, evar. But please dont, cuz Im mining right now. Let you know when Im done kthxbai.

[Vladimir]

AND, this is the first time I've seen the block average at bitcoincharts.com under 6/hr!

CRAY was testing their new supercomputer on the Bitcoin network and realized they were losing money because it has nVidia GPUs?  

LOL! thank you for this one.

[Vladimir]

It got up to 86 here in New Hampshire the past few days. It's HOT.

Suddenly, mining operation in UK where average temperature is 10 C and rarely lover that 0 C in the winter plus lots of cold winds does not sounds as such a bad idea.

[commlinx]

I got scammed by this too, recently I bought one of the cheap ASIC miners on ebay that worked great for a day or two then I got an e-mail it would stop working at block 127225 unless I paid a further $1000 per core for an unlock code. I refused the extortion and guess many others were the same. At 45 Ghash/s for 16 cores it might have been worthwhile but I couldn't be sure I wouldn't be scammed again at some future block count

huh?

yes, please elaborate...

Nevermind the guy had his ebay and Paypal account suspended. I offered one of my mates who breaks DRM and copy protection for The Scene 500BTC if he could crack the problem. He'd never heard of Bitcoins before but loves that sort of challenge and was keen learn more about Bitcoin as well. He disassembled the client and found the unlock code was just a simple XOR type thing between the cutoff block and a static block in the code, along with a simple CRC32 check to validate the code and a few seconds time delay to make brute-forcing impractical.

Anyway just firing it all back up now that I've got the miner back from him, so soon hashrate should be back up to normal .

[dikidera]

Well, it was originally 1,8THash/s when i joined a month ago.

[borito4]

aaaaand its on its way up again