Bitcoin difficulty may actually go down in ~7 days!

[philipma1957]

and 3.08%  jump just happened.   So put 3% in your calculators and see how good  you miner looks.

[TheJuice]

bitcoinwisdoms diff algos rely heavily of the previous change. Thus early on in an adjustment period it is basically saying the diff will inc by the prev amount. Right now my models show 2% increase.

My models rock.

Next difficulty looks like 7-9% based on expected delivery.

[kendog77]

I'll gladly take a 3% difficulty increase to offset that awful 25% increase we just had!

Unfortunately, I think S3s will start rolling off the assembly line by the hundreds so the next difficulty increase will probably between 10-15%.

[ALToids]

I'll gladly take a 3% difficulty increase to offset that awful 25% increase we just had!

Unfortunately, I think S3s will start rolling off the assembly line by the hundreds so the next difficulty increase will probably between 10-15%.

Depends on what day of the week they ship out and if miners can get them online early enough into this difficulty.  Right now doesn't appear any significant change has occurred since it jumped the 3%.

[davejh]

I'll gladly take a 3% difficulty increase to offset that awful 25% increase we just had!

Unfortunately, I think S3s will start rolling off the assembly line by the hundreds so the next difficulty increase will probably between 10-15%.

Depends on what day of the week they ship out and if miners can get them online early enough into this difficulty.  Right now doesn't appear any significant change has occurred since it jumped the 3%.

Unfortunately there's very little that we can infer from the hashrate immediately at the difficulty changes. Over the last year the instantaneous numbers can be overshooting or undershooting at that point. I've been working on an assumption that 13% to 14% is the "norm" and that the 25% was an extreme overshoot. The 3% we've just had is really putting things back to where I'd have expected. By the end of the year I expect we'll be under 10% per diff change but difficulties will inevitably be erratic just by virtue of the way mining stats work.

[kgb2mining]

if the s-3 can under clock / under volt like the s-1 it will be at .5-.6 watts  .

 say 400gh at 200-240 watt.  it will be better then a sp30

Even at those numbers, they're close,  but the sp30 still pulls ahead by a slight margin:

400 Gh/s / 240watt = .6  or 400 / 200 = .5
6 Th/s / 2400 watt = .4

Also price-wise very close (hardware purchase only):

11 x S3 = $5027.91 (at $630/1BTC today) / 5.357 Th/s (@stock 487 Gh/s, @550 if possible it evens it up)
$5,095 / 6 Th/s

Stock voltage:
11 x S3 = $5028 / 5.357 Ths / 4Kw
1 SP30 = $5095 / 6Ths / 2.4Kw

Undervolt assuming best case 400/200:

11 x S3 = $5028 / 4.4Ths / 2.2Kw
1 SP30 = $5095 / 6Ths / 2.4Kw

Now, the one thing that's going in the S3's favor is the form factor.  The SP30 is harder for the home hasher to deploy since it's rackmount and it's going to be loud as all heck with all those fans.  For us in a datacenter setting it makes more sense for the rackmount as our S1 deployment is nice but kludgey.

For today tho, it's still hard to beat used S1's if you have cheap power.  Our average cost has been about $160 each, and they're very stable overclocked @205Ghs, so we have roughly 4.7Ths deployed for about $3700.  Only after about 90 days it evens up and the S3's start to pull ahead counting power.

[ShakyhandsBTCer]

I'll gladly take a 3% difficulty increase to offset that awful 25% increase we just had!

Unfortunately, I think S3s will start rolling off the assembly line by the hundreds so the next difficulty increase will probably between 10-15%.

Depends on what day of the week they ship out and if miners can get them online early enough into this difficulty.  Right now doesn't appear any significant change has occurred since it jumped the 3%.

Unfortunately there's very little that we can infer from the hashrate immediately at the difficulty changes. Over the last year the instantaneous numbers can be overshooting or undershooting at that point. I've been working on an assumption that 13% to 14% is the "norm" and that the 25% was an extreme overshoot. The 3% we've just had is really putting things back to where I'd have expected. By the end of the year I expect we'll be under 10% per diff change but difficulties will inevitably be erratic just by virtue of the way mining stats work.

If you take the average of the 3% increase and the previous 25% increase then the total increase for the two periods is roughly the same as if the difficulty increased 14% both periods.

[lightfoot]

Pretty much. Mining went from 30% increases to 15%. Which stretched out the operational time of my singles for a month or two. Not sure where things will go from here, if the price of bitcoin does not rise, then we will hit a technical limit as to how much money people want to donate to the blockchain (ie: none IMO).

Which will stabilize mining at a level where people can grind out a small profit. Which should be replaced by the damn transaction fees.

[philipma1957]

Pretty much. Mining went from 30% increases to 15%. Which stretched out the operational time of my singles for a month or two. Not sure where things will go from here, if the price of bitcoin does not rise, then we will hit a technical limit as to how much money people want to donate to the blockchain (ie: none IMO).

Which will stabilize mining at a level where people can grind out a small profit. Which should be replaced by the damn transaction fees.

yeah, asics are now common.   june 2013 to dec 2013  diff was 24%   jan 2014 to july diff was 18%

 the power efficiency  jumps are less and diff has shrunk a bit.


When I was mining gpu's  back in the day they would grind out a small profit.

 asics are drifting into this.  we will see as time unfolds.

[RoadStress]

Now, the one thing that's going in the S3's favor is the form factor.  The SP30 is harder for the home hasher to deploy since it's rackmount and it's going to be loud as all heck with all those fans.  For us in a datacenter setting it makes more sense for the rackmount as our S1 deployment is nice but kludgey.

SP30 will be less noisy than the SP10, but yes the datacenter setting is preferred for various reasons.

[taipo]

Hows the hashrate/diff increase looking for this next round?

[freedomno1]

Hows the hashrate/diff increase looking for this next round?

Not that bad actually the last round went up around 8% and the round before that 3% so its quiet compared to before for a while
The hashrate is increasing but not rapidly at this time looking like a flat round or small increase again for the next round
The last increase was 2 days ago though so its still early in the cycle.
Between 0 to 5% perhaps
http://bitcoindifficulty.com/
https://bitcoinwisdom.com/bitcoin/difficulty

Jul 25 2014   18,736,441,558   8.08%   134,120,673 GH/s
Jul 12 2014   17,336,316,979   3.08%   124,098,191 GH/s
Jun 29 2014   16,818,461,371   24.93%   120,391,236 GH/s

[philipma1957]

yeah I agree 5 to 10 % this time.

one reason is a lot of gear above 2 watts is being pulled off the market.

the other reason is most miners are doing summer not winter.  so heat cuts down mining.

another reason is new gear is not .3 or .2 watts so it sells but not as fast.

[freedomno1]

bitcoinwisdoms diff algos rely heavily of the previous change. Thus early on in an adjustment period it is basically saying the diff will inc by the prev amount. Right now my models show 2% increase.

My models rock.

Next difficulty looks like 7-9% based on expected delivery.

Ah just noticed someone who got that last one spot on 8 was in the middle lol.
Guess I should ask TheJuice when they get back on what they predict

@ Philipma

True enough summer heat makes mining more costly so we might see the rate boom in the winter months and stall for the next few until it gets colder.

[philipma1957]

I get these two


http://bitcoincharts.com/                                  next diff = 18701.6 with 1836 blocks to go


https://bitcoinwisdom.com/bitcoin/difficulty      next  diff = 19644.3   with 1838 block to go


add them divide them  come up with   19172.9     next jump  which is 2.32%  over current 18736.4


I do this each day  and make my moves based on this and btc prices.


  Diff has show long term drops if you do 200 day periods

dec 21 2103  to july  12 2014 is just about 18%

jun 5 2013 to dec 21 2013  is just about 25%


Simple reasons are asics fully  replaced fpga and gpu   in the jun to dec time of 2013  this meant huge growth due to really good power savings.

from dec 2013 to now 2 watt s-1 becomes .75 watt s-3 and that is  less growth due to okay power savings.

 btc price drop from dec to now also has slowed growth.


next 200 days  .75 watt drops to .4 watts?  = slower growth.

btc price can jack up growth and  maybe someone invents a better way to hash.. and new gear  goes  much faster with better power.


My guess is no. 

 look at gear like a hard drive :


 ssds'   were big jump over hdds' and now we have flattened out in   drive speed increase.

 With  the new  ssds 2x the speed of older ones.  not quite the same but you get the idea.

[freedomno1]

I get these two


http://bitcoincharts.com/                                  next diff = 18701.6 with 1836 blocks to go
https://bitcoinwisdom.com/bitcoin/difficulty      next  diff = 19644.3   with 1838 block to go
add them divide them  come up with   19172.9     next jump  which is 2.32%  over current 18736.4

I do this each day  and make my moves based on this and btc prices.

  Diff has show long term drops if you do 200 day periods
dec 21 2103  to july  12 2014 is just about 18%
jun 5 2013 to dec 21 2013  is just about 25%

Simple reasons are asics fully  replaced fpga and gpu   in the jun to dec time of 2013  this meant huge growth due to really good power savings.
from dec 2013 to now 2 watt s-1 becomes .75 watt s-3 and that is  less growth due to okay power savings.

 btc price drop from dec to now also has slowed growth.
next 200 days  .75 watt drops to .4 watts?  = slower growth.

btc price can jack up growth and  maybe someone invents a better way to hash.. and new gear  goes  much faster with better power.
My guess is no.  

 look at gear like a hard drive :
 ssds'   were big jump over hdds' and now we have flattened out in   drive speed increase.
 With  the new  ssds 2x the speed of older ones.  not quite the same but you get the idea.

Thanks I do the same with those charts but just look at their calculation they estimate it at on a daily basis your internal average would put it down to a far narrower measure and be more accurate over longer duration since trends would start to appear.

And your right FPGA's and GPUs are long out of the Bitcoin market and moved to altcoin mining which oddly enough is slowly getting replaced by ASICs as well so the areas of focus are increasing chip concentration 32 mm to 20mm etc and efficiency per watt hour.

Back to Bitcoin the centralization of mining centers is an interesting development that I am watching as the single units become less commercially viable options.
I'm surprised that liquid cooling hasn't yet emerged in larger scales though but that would probably extend the lifespan of mining units over the summer by keeping the heat costs down.

https://asktom.cf/index.php?topic=346134.0
http://v.youku.com/v_show/id_XNjk5MTc0MTY0.html
AM has a few of those facilities up so considering the scale increasing as larger commercializations of these central units start to emerge I expect this variance to more or less fade away in a year or two but your right at present the difficulty will slow down and decrease in jumps for a while.

Thanks that was interesting stuff and likely explains this years trendline, the second part I wrote probably covers 2015 to 2016 summer deployments. Chip to Unit direct mobilization.

[taipo]

The clue here as to whether there is a new jump in tech in the wings is that when Bitcoin kicked off there was CPU mining, the potential of GPU mining, FPGA and ASIC miners could also possibly be developed. That is all complete now, with advances of late being improvements in power consumption.

The last thing waiting in the wings is the mythical unicorn of hashing, that of the quantum computer. But that is not on the near horizon.

So is this Peak Hardware we are heading to in the next couple of months?

[philipma1957]

The clue here as to whether there is a new jump in tech in the wings is that when Bitcoin kicked off there was CPU mining, the potential of GPU mining, FPGA and ASIC miners could also possibly be developed. That is all complete now, with advances of late being improvements in power consumption.

The last thing waiting in the wings is the mythical unicorn of hashing, that of the quantum computer. But that is not on the near horizon.

So is this Peak Hardware we are heading to in the next couple of months?

maybe  but as the Northern hemisphere   heads to fall /winter some offline gear  will come back online.  say 5-10% at a minimum

[klondike_bar]

The clue here as to whether there is a new jump in tech in the wings is that when Bitcoin kicked off there was CPU mining, the potential of GPU mining, FPGA and ASIC miners could also possibly be developed. That is all complete now, with advances of late being improvements in power consumption.

The last thing waiting in the wings is the mythical unicorn of hashing, that of the quantum computer. But that is not on the near horizon.

So is this Peak Hardware we are heading to in the next couple of months?

pretty much. over the next 6 months all manufacturers will be using 20nm tech probably, and undervolting for efficiency, to a peak of maybe 0.25w/GH

[TheJuice]

bitcoinwisdoms diff algos rely heavily of the previous change. Thus early on in an adjustment period it is basically saying the diff will inc by the prev amount. Right now my models show 2% increase.

My models rock.

Next difficulty looks like 7-9% based on expected delivery.

Actual increase was 8.08%. Rock on new models.

Next predicted difficulty increase: 4-6%