Bitcoin Forum

I've create a spreadsheet to roughly estimate the coming mining difficulty.

It estimates the anticipated hash rate of the network each month until December. I don't realistically expect anyone could have visibility past december. That being said, after December, the spreadsheet calculates a flat 10% growth per month.

The idea is to determine what the break even cost (B/E) for a mining unit might be.

For example, if you could obtain a 66 Gh/s unit starting in June, your 6 month profit would be estimated at 83 btc. Therefore, as long as you pay under 83 btc for the unit you will make a profit in 6 months.

However, if you were to acquire the same unit in October, then your B/E price over 6 months would be 25 btc.

I've chosen 6 months as the timeframe because I believe there is simply too much risk / variability past 6 months to accurately calculate a ROI.

The spreadsheet calculates the anticipated hash rate, shows an anticipated network difficulty, and then calculates the estimated btc / day rate based on the network hash rate.

I would welcome any discussion as to the formulas in the spreadsheet and the assumptions underlying when each company would bring their hash rate to the network.

The spreadsheet can be accessed here (https://docs.google.com/a/qubitpool.com/spreadsheet/ccc?key=0Al6CLYo1WC9VdGYzaGxOMndBbEZ4YldqTG93eV9GM1E#gid=0).

Cheers!

Cool!

Your spreadsheet is so full of errors it's unbelievable. Avalons Batch 2 and 3 will have approx. 43th each (72gh x 600 units, and that is with a +10gh bonus for a margin). not 156th or 130th. Where the heck did you get that numbers?
Also: The hell?!? Are you really thinking that BFL can pull 58th per month out of their hats? That would be a great magic trick! Besides, I think if they have some th's ready, they are mining for themselves right now, so they would be in the current number.
Then there is the Asic Chips: .... There are over 500.000 chips ordered, why do you add them in with only 85th? That should be a number between 150-200th. If all of them are okay and up and running in a ridiculously short amount of time.

Aaaaand: Unknown 500th?!?  Where do they magically appear?

If you make a spreadsheet, at least get your numbers straight!!!   (Knc miner numbers are also wrong. I alone have preordered over 1 th in mining equipment from them... If they deliver...)

yeah, his calculation is about 4x of the actual ghs.

I estimated at the end of the year, difficulties is about 50m-100m. Currently 12m.

The reason I posted the spreadsheet is for help in vetting the numbers with the community. I appreciate your constructive feedback. My plan is to update the spreadsheet each month with new information as it appears. You will find the information you've provided on the "July" spreadsheet.

You are correct avalon batch 2 & 3 are 600 units representing 43Th.

For the avalon chips, I've updated the information based on the btc received, estimating the total chips at 760,000 and the effect as 215 Th.

As a conservative estimate, for now, I'll leave the BFL prediction, although I may reduce it later this month based on the assumption that BFL chip sales indicate an inability to deliver units.

The unknown factor represents what I anticipate to be asic projects that are privately in development that will not be announced to the public but will affect the network hash rate.

Also, for knc miner, the November / December figures represent delivered (not ordered) units. Again, just an early estimate for now.

Cheers!

Better change the title of this thread, it's misleading.

My idea was to create a spreadsheet that would be updated each month with the latest information. That information would come from the community and ideally would result in the prediction changing as new information was added or changed.

Thus, the spreadsheet would serve as a guide to help estimate the profitability of equipment.

The change I've made so far have resulted in the December estimate decreasing from 252 million to 226 million.

I'd rather have the title of the thread reflect a stick in the ground when I first made the post rather than practicing revisionist history. In the mean time the spreadsheet will reflect the latest thinking.

Then we'll see in December how close it was  ;D

while 250,000,000 might be a stretch, 100,000,000 certainly is not come December. This is still an incredibly scary number  :o

I did all my calculations with 100mil as the worst diff for October. Most miners still get a lot of cash out of them at that point. If I remember correctly even a 60gh device gives you more than 1k a year at 250 mil...

Hi guys, if u all interested in difficulty, why not make a profit by guessing future difficulty? I sincerely invite you guys to come and check out this website.  http://btcfarseer.com (http://btcfarseer.com)

I did an estimate of 230 million in Dec 2013?

Optimator, thanks for the spreadsheet. Great initiative!

Tinoki, are you stating? Can you post your numbers?

It's been an exciting week of announcements. I'll start to aggregate the additional information later this week. Information from KNC, bitfury, and, it's small, but there are also Avalon clones (https://bitcointalk.org/index.php?topic=215431.0) coming to market.

I'll probably add these hashes while removing them from the "unknown".

MY personal bet is around 60-70 million, because of the hardware costs.

Taking BFLs price of a 50gh miner as a price point, which is the lowest price point there is at the moment, I suggest that an increase of 10 million actually costs about ~4 million dollars at the moment.

An increase to 250 million would therefore come with a buyin cost of around 250 million dollars.

Considering the amount of money people have spent on the smaller ASIC miners, FPGAs and preorders already, I would argue that for some time, this market will stall. Basically, at this point, unless another April 16 spike hits with thousands of newcomers, this market is saturated for some time with people paying absolute premiums for mining equipment.

I guess that 60 million, with an additional cost of 24 million dollars at BFL's price point is realistic until the end of the year. I might take 70-80 million, which would indicate an investment of around 32 additional millions of dollars.

But 250 is too heavy a price to take. Before generation 2 chips enter the market and the Bitcoin reaches somewhere around a STABLE 250, I doubt we will see anything beyond 120-130 million. Should both scenarios occur however, it is gonna become the next arms race.

It is a good idea to check back the actual costs of additional mining power. you will have ASIC miner ruin this calculation a little, since they have a $10.000 per added Terahash price point, but fortunately they are NOT giving this price to people yet.

hmm I should be shorting bitcoins then


with the size of blocks average 25.20 btc one can clearly see that more coins are being added than # of transactions to justify the size of the economy

the difficulty jump today is expected to be 28% so your spreadsheet makes too many assumptions

Excellent points! I agree that at some point the market will stall until either a new gen of chips appear or the price increases. I've been trying to figure out how to calculate the "stalling point".

If we take your figure of 60 Million as the difficulty and 12 million as the current difficulty the net increase is 48 million. 48 million represents about 80% of the 60 million network hash.

To gain 80% of the network $24,000,000 was paid. At $120 / btc the network generates $432,000 per day in block reward. Investing $24M your payback would occur in about 55 days.

It seems people would still invest at that point??


Can you explain this a little more?

In short terms: If you want to add hashing power, someone has to build it and someone has to pay for it. In USD, Yuan, Rubels or Euro.
So if you add hashing power to the network, someone needs to have spent this money. That is why a calculation that simply adds 18% every month is unlikely. At some point, a linear increase is going to have such significant costs that it becomes unlikely, even if your hasing power becomes ridiculously bigger through new chips.

I believe this is the point where not only mining, but also running full nodes should be valued by people. Getting paid for a) mining and b) confirming as a node will become very important for stability and security.

Nice work :)
Does it include this latest info?

http://thegenesisblock.com/latest-shipment-of-avalon-asics-could-increase-network-hashrate-by-500/

Woof.

While Avalon has not released information how many chips have been ordered, a 135 box shipment would amount to over 2.1 million ASIC chips. Given that Avalon sells them in batches of 10,000 for 780 BTC, this lot alone is worth over 164,000 BTC ($14,700,000).

The Kaiju have arrived.
http://thegenesisblock.com/wp-content/uploads/2013/07/avalon8.jpg

I haven't updated the End of July estimates with the latest info. I'll revisit the chips and I'm also looking at adding AMC / Bitfunder.

In order for your prediction to be true:

You are predicting 4,238,893 GH/s in 12 months(592 Million Difficulty).  Lets looks at what that means.

1) This says there will be ~ 70,000 units mining at an average of 60 GH/s in 12 months....
2) That would be 6000 units per month shipping starting right now and sustaining for 12 months....
3) That equates to a minimum of $18,000,000 per month being spent in hardware every month for the next 12 months (at current prices)...
4) That equates to $3,204,603.108 / month being spent on power if you consider usage at ~7w / GH @ $0.15 / kwh
5) The amount of coins currently being generated in 30 days is worth $10,800,000 at current prices.
6) That means that $254,455,236 would have been spent to generate $129,600,000 worth of value.

I highly doubt that the equation will become that skewed.  The market is mostly rational and people will stop "investing" in asics long before we hit this difficulty level. 

Or 14,000 units hashing at 300GH/s. Larger units than that for less money than a Batch 3 Avalon are on the horizon.
It is not out of the realm to think that 1200 such units could ship per month over the entire Bitcoin ecosystem.
5,30 & 60 GH/s are first generation sized ASIC mining devices. Density will increase and cost will decrease.

~11,000 Jupitors from KNC could supply that much hashrate.
At their current prices, that many Jupitors would cost $75,900,000. That hash rate would be chasing after $129,600,000 worth of value.
Obviously, future purchases of mining equipment would be spread over many vendors.

Your point is valid. While I continue to doubt any rational market hypothesis, I agree that there will be an equilibrium point. The primary agent working against a hard equilibrium point is btc price speculation. There may well be individuals or even companies who invest in mining on price rise speculation.

I do believe 252,000,000 overestimates the December difficulty, however, there will be generation 2 and generation 3 chips coming to market. I'd estimate generation 2 chips within 9-12 months and gen 3 chips in 12-18 months.

As the new chips hit the market the cost per gigahash will decrease. Using US dollars as a baseline, Avalon batch 1 units cost about $18 / Gh. KNC Jupiters are at about $17.50 / Gh. Butterfly Labs is around $50-$55 / Gh, although their initial pricing was closer to $18 / Gh. Gen 2 chips may reduce that cost by 75% and gen 3 chips by 50%.

The point being that more hashrate can be brought to the network at substantially reduced costs while speculators may invest counting on btc to increase.

Cheers!

To create such things, at this moment, I currently estimate how bad when the difficulity is 60M based with current method known. The answer, well, I think at the december 2013, About 55% of your estimated as Avalon cannot stand last, The butterfly almost do nothing, Knc is questionable, the ASICMiner do very well.

From kncminer currerent status, I currently have little idea how many modules will be made by ORSoc based on order number.

I think, for preallocated 160-180Thash/s mentioned, there are at least 4000 modules will be made, with different configurations.

To create the number, I think you will need 20K modules to do.

Interesting spreadsheet.

So, by the end of December 2013, is it even worth it to GPU mine for Bitcoins anymore?  Hmmm.....

As always, mining is dependant on power cost. If you don't pay for power, GPU will always be worth it.

Clearly wrong. July is almost over and we just hit 31 million, which is 23 million lower than your spreadsheet suggested. I say it's possible we hit 125 million by the end of the year, but 250 at this rate doesn't seem likely.

Any update guys on what December may look like? I don't see 250 million, even with all the new players. I get the feeling they will be late to deliver. Look at the recent Avalon debacle/scam...

I'm guessing 120-150 million in December. Ideas? Links?

Thx,
IAS

Here is a Wolfram Alpha analysis of that last few retargets fitted to an exponential:
http://www.wolframalpha.com/input/?i=exponential+fit+%7B%7B13%2C+3651012%7D%2C%7B37%2C+4847647%7D%2C%7B96%2C+11187257%7D%2C+%7B120%2C+15605633%7D%2C%7B156%2C26162876%7D%2C%7B167%2C31256960%7D%2C%28189%2C50810340%7D%7D

I have complete faith that 100TH mine project will deploy 200TH mine in September.  Avalon chips will ship and DIY folks will run their miners so their losses will be smaller.https://bitcointalk.org/index.php?topic=236050.0

If the curve follows the exponential (which I fully anticipate it will)  we should 500M difficulty Jan.1 as a conservative estimate, and 1B difficulty as a stretch.

Holy Crap! Nice work there but I do hope your numbers are a bit off. If all these companies ship and your conservative exponential growth is right on, I guess it is possible.

Thanks a lot for posting this. I guess we'll watch what happens in September, as that is when some units begin shipping. October is where the shit might hit the fan...