I Predict Difficulty Will Be Over 100 Million by Summer 2013

[Obsi]

This is mostly for my own amusement. Bump this to tell me how wrong or right I was when summer arrives 

[Korbman]

This is mostly for my own amusement. Bump this to tell me how wrong or right I was when summer arrives 

So..you're saying a 33x increase by summer? That would require a network hashrate of 726TH/s.

Let's take BFL's current preorder rate of 31TH/s...let's say double that for people that haven't put in their orders on the forum. So now 62TH/s.
Meh, let's double it again and assume that the other ASIC companies out there have just as many preorders combined to make a similar rate. Now we're up to 124TH/s.

This 124TH/s is going to be our base value for Q1 2013...you're predicting that the number of total orders (and overall devices in the wild) would increase 5.85 times within the next 6 months of Q1. The only way that would happen is if the following are met:
1) ASIC devices not only meet their predicted specs, but some exceed them
2) ASIC devices become extremely cheap
3) USD/BTC values skyrocket to new highs and stabilize up there

Only if all three of those conditions are met do I think the difficulty would reach that high...mainly since everyone and their grandma who even heard about bitcoin in passing would be trying to make some easy money with cheap ASIC devices.

[ercolinux]

Maybe not 100.000.000 but neither only 17.000.000 as Korban say.
I extimate a 3-40.000.000 difficulty in 8-10 months: there are ASIC for 1THash/sec in delivery for the Q1 of 2013, but is not impossible to expect some 2 or 3THash/sec rig in near future.
The Korban prevision is based on the assumption that no one buy more ASIC after the first batch (or near so), but I think that for every ASIC preorderd at least another one can be sold in the next 4-6 months, doubing the difficulty.

[scrybe]

I've got my eye on 58 million by 7/5/13, so your 100M would be less than twice as much. (edit: I kept reading double as doable)

I doubt we will see more than 70M, but 100M is almost a certainty by the end of 2013.

Current Prognostication from my optimistic model:
Block#   Date & Time   Block Time      Block Value      TimeToFind      2016 time         Global Payout $   Global Hashrate   Difficulty       USD/BTC Rate    Earnings per GH/day
245952   7/5/13 6:10      9.642791286   25         1166392.034   13.4999078   49000   $776,133.19    414641       57,924,637.73     $15.84       0.012155102

I do think we are going to see a bit of a bubble towards the end of the year, and a slump before growth returns in the spring or early summer.
Some implications of this:

• More GPU's than expected are going to be forced from the market if the hasing rate is up in the 30 million range while BTC drops to the $4-8 range somewhere
• Some FPGA's may even get turned off for a while.
• Those expecting to "re-up" every 3-4 weeks as they make mining profits will see their timetables go to hell, or switch to saving

(oh, that is my Median Risk sheet BTW, I do have a model that hits 200M by E0Y 2013, but it is based on assumptions like 100% of BFL pre-orders numbers shipping, which are a bit unreasonable.


Well, that is my guess.
Someone witness this baby and we will see if it is ugly in 9 months.

[Korbman]

Maybe not 100.000.000 but neither only 17.000.000 as Korban say.
I extimate a 3-40.000.000 difficulty in 8-10 months: there are ASIC for 1THash/sec in delivery for the Q1 of 2013, but is not impossible to expect some 2 or 3THash/sec rig in near future.
The Korban prevision is based on the assumption that no one buy more ASIC after the first batch (or near so), but I think that for every ASIC preorderd at least another one can be sold in the next 4-6 months, doubing the difficulty.

How did you get those numbers?...and no where did I say anything 'no more purchases after the first batch'. I was just making numbers up for this example.

My actual research puts the difficulty at around 30M over the course of 12 months...but that's about to be updated given BFL's new product news, as well as increased ASIC competition and more preorders.

[Obsi]

This is mostly for my own amusement. Bump this to tell me how wrong or right I was when summer arrives 

So..you're saying a 33x increase by summer? That would require a network hashrate of 726TH/s.

Let's take BFL's current preorder rate of 31TH/s...let's say double that for people that haven't put in their orders on the forum. So now 62TH/s.
Meh, let's double it again and assume that the other ASIC companies out there have just as many preorders combined to make a similar rate. Now we're up to 124TH/s.

This 124TH/s is going to be our base value for Q1 2013...you're predicting that the number of total orders (and overall devices in the wild) would increase 5.85 times within the next 6 months of Q1. The only way that would happen is if the following are met:
1) ASIC devices not only meet their predicted specs, but some exceed them
2) ASIC devices become extremely cheap
3) USD/BTC values skyrocket to new highs and stabilize up there

Only if all three of those conditions are met do I think the difficulty would reach that high...mainly since everyone and their grandma who even heard about bitcoin in passing would be trying to make some easy money with cheap ASIC devices.

I think we'll be in the PH/s range sometime next summer. BFL's bitpay order numbers are rising at ~81.5 per day. Assuming the orders average out to ~60 Ghash per order (which I don't think is too far off base with their current released specs, since there are multiple claimed SC rig orders, and orders of 6+ singles for one order number) that's ~5 Thash network growth per day just from BFL. Granted, many of those orders are probably not completed & paid for, but that is just the BitPay payment method we're looking at. People can also order through PayPal & Bank Wire.

Then you have 4-5 competitors releasing ASIC products as well. We've already seen a bit of competition driving the $/Ghash down, I think we'll see much more of that in the coming months. Some people have claimed that the chips can be produced for less than $1 per Ghash. If that is the case, I don't see how we won't have a multi-petahash network speed next year.

Of course this argument goes out the window if no one delivers on spec, but it will be interesting times for miners either way.

With the reward halving and coming ASIC price wars we might see first gen ASIC miners with high electricity rates priced out of the market sometime in 2013.

[ercolinux]

How did you get those numbers?...and no where did I say anything 'no more purchases after the first batch'. I was just making numbers up for this example.

My actual research puts the difficulty at around 30M over the course of 12 months...but that's about to be updated given BFL's new product news, as well as increased ASIC competition and more preorders.

I think 30M will be reached in 8-9 months (unless more delay in ASIC production)
You consider only the preorder of BFL + a similar batch from other vendors and double it for security: total 124Thash. But Avalon has a preorder of 300 60GHash rigs alone for feb.2013 (18GHash) and say that another similar (or even bigger) batch will be ready after that date in few months, 62Thash will be from BFL and we are at 100THash by Apr-may. There are at least other 2 or 3 ASIC announced, if they will sell 60-80Thash in 6 months we are at 160-180GHash. Some FPGA and GPU miner still continue the operations due to free current or similar reason and we have reached the 30M.

[Korbman]

I think 30M will be reached in 8-9 months (unless more delay in ASIC production)

Well how about that..we're on the same page.

You consider only the preorder of BFL + a similar batch from other vendors and double it for security: total 124Thash.

Nope. Made up those numbers off the top of my head for an example.

But Avalon has a preorder of 300 60GHash rigs alone for feb.2013 (18GHash) and say that another similar (or even bigger) batch will be ready after that date in few months, 62Thash will be from BFL and we are at 100THash by Apr-may. There are at least other 2 or 3 ASIC announced, if they will sell 60-80Thash in 6 months we are at 160-180GHash.

Link to this?
And yes that seems about right. My actual prediction starts at 150TH/s and goes up to 250TH/s within a 12 month timeframe.

[Korbman]

I think we'll be in the PH/s range sometime next summer. BFL's bitpay order numbers are rising at ~81.5 per day. Assuming the orders average out to ~60 Ghash per order (which I don't think is too far off base with their current released specs, since there are multiple claimed SC rig orders, and orders of 6+ singles for one order number) that's ~5 Thash network growth per day just from BFL. Granted, many of those orders are probably not completed & paid for, but that is just the BitPay payment method we're looking at. People can also order through PayPal & Bank Wire.

Yeah, that's not quite likely. I alone have put in dozens of fake orders alone to test out different combinations of equipment. Turns out there are plenty of people out there like yourself that think I actually ordered all of them just because the order number incremented by 12.

Then you have 4-5 competitors releasing ASIC products as well. We've already seen a bit of competition driving the $/Ghash down, I think we'll see much more of that in the coming months. Some people have claimed that the chips can be produced for less than $1 per Ghash. If that is the case, I don't see how we won't have a multi-petahash network speed next year.

As impressive as that would be, it would also be the only way to achieve the PH/s mark you noted. At current prices, a 1PH/s network rate would mean you're $30,000 Mini Rig only churns out 200BTC a month. Granted that would mean about a 12 month payback (at the current rate of $12.35 USD/BTC), which isn't terrible. Buuttttt, if that rig price would drop to $3,000 or less, then I'd say your prediction is right on track.

Of course this argument goes out the window if no one delivers on spec, but it will be interesting times for miners either way.

True, but my guess is that it's pretty accurate. There are multiple companies offering roughly the same hash rate per chip...so either they're all conspiring against us or the chips do have some potential for high, sustainable rates.

[Obsi]

Yeah, that's not quite likely. I alone have put in dozens of fake orders alone to test out different combinations of equipment. Turns out there are plenty of people out there like yourself that think I actually ordered all of them just because the order number incremented by 12.

A single SC minirig order easily absorbs your fake/unpaid orders while still boosting the average order size.

Look at the stats here: https://bitcointalk.org/index.php?topic=89685.0

227 tracked orders

225 Jalapeno SC            
231 Single SC
21   Mini Rig SC

Total Hashrate on order from tracked "valid" orders: 46.3725 Thash

That's 200 Ghash per valid order. Do you think fake orders exceed real orders enough to bring that average down below 60 Ghash?

Assuming the average stays the same for unreported orders (which is a huge assumption I admit) and assuming half of all orders are fake, that means with 7678 orders in the pipeline, at 100 Ghash per order due to half being fake... there is 767.8 Thash just from one payment option from just one manufacturer.


Edit: Also note that this means the future network hashrate is growing at 4-5 Thash per day right now from this one order pipeline. So as BFL ramps up production and catches up to all these preorders we might see a very fast explosion in network speed coming over a timeframe of just a few months. Placing an order for hardware right now means once you receive it the network speed may very well exceed 1 Phash. Worst case scenario profitability calculations should take these possible factors into account.

[Korbman]

Total Hashrate on order from tracked "valid" orders: 46.3725 Thash

That's 200 Ghash per valid order. Do you think fake orders exceed real orders enough to bring that average down below 60 Ghash?

Assuming the average stays the same for unreported orders (which is a huge assumption I admit) and assuming half of all orders are fake, that means with 7678 orders in the pipeline, at 100 Ghash per order due to half being fake... there is 767.8 Thash just from one payment option from just one manufacturer.

Right, but those are also orders over the past 3 months. Granted with the updated news I'm sure more valid orders are going to go in (probably a lot more), but I just can't picture anything near 770TH/s..even when adding up all of the competition as well.

We'll just have to see where things go over the next 3 months. Maybe we'll have a better idea of the direction around then and we can speculate half as much

[MrTeal]

750TH/s is insane, and we won't be there by summer 2013. To get that hashrate using Minirig SC's, you'd need 500 of them or $15M worth of equipment. The market is getting bigger, but I don't think that there's going to be that much new equipment coming online in a half a year.

[Obsi]

Lets look at it from a different angle... from a quick count of FPGA singles orders that were tracked here:
https://bitcointalk.org/index.php?topic=77796.0

There appears to be 568. Now no one in his right mind would pass up the chance to upgrade those singles to ASICs. Either they will sell them to someone who intends to upgrade them or buy the upgrade themselves. The upgrade opportunity ends in March 2013 I believe. So what portion of FPGA singles are on that list? 10%? 50%? 90%?

So that portion of singles which have been tracked represents future hashing ability of 60 Ghash each.

568*60 Ghash = 34.08 Thash

What portion of all FPGA singles are tracked?
90%? = 37 Thash minimum network growth from BFL FPGA single upgrades alone
50%? = 68 Thash minimum network growth from BFL FPGA single upgrades alone
10%? = 340 Thash minimum network growth from BFL FPGA single upgrades alone

In addition, we've got mini-rig upgrades, flat out purchases, people just jumping in with a jally that never thought of mining before. The new SC single "lite" that is planned at 30 Ghash plus 4 or 5 other manufacturers filling up their preorder queues right now.

[Graet]

summer where?
Bitcoin is a lot international - for me summer is weeks away.....I cant see diff being that high in a few weeks

[hahahafr]

In the next six months difficulty will gradually increase to 34x. After that the 2nd generation of ASICs will arrive and nobody knows how they gonna perform.

[Obsi]

summer where?
Bitcoin is a lot international - for me summer is weeks away.....I cant see diff being that high in a few weeks

To clarify, around June-July 2013.

[Korbman]

In the next six months difficulty will gradually increase to 34x. After that the 2nd generation of ASICs will arrive and nobody knows how they gonna perform.

Wrong. In the next six months the difficulty will increase 475x.

Oh I'm sorry, I thought we were pulling random numbers out of our asses.

[hahahafr]

In the next six months difficulty will gradually increase to 34x. After that the 2nd generation of ASICs will arrive and nobody knows how they gonna perform.

Wrong. In the next six months the difficulty will increase 475x.

Oh I'm sorry, I thought we were pulling random numbers out of our asses.

https://bitcointalk.org/index.php?topic=110138.msg1234682#msg1234682

[Sitarow]

In the next six months difficulty will gradually increase to 34x. After that the 2nd generation of ASICs will arrive and nobody knows how they gonna perform.

Wrong. In the next six months the difficulty will increase 475x.

Oh I'm sorry, I thought we were pulling random numbers out of our asses.

Then lets do the 4x max per difficulty cycle and assume that there will be 3 difficulty cyles within two 4 weeks from shipping.

If 30 of the estimated 46.3 TH/s from BFL source go live, that puts the  Network total to 53.144 Thash/s

At 23.2Thash/s the Estimated    difficulty will be 3,060,648 in 279 blks

Add 43.6% to the Estimated Difficulty.

At 53.2Thash/s the Estimated difficulty will be 4,395,090 add 6.5% for every 2 weeks that those 30Thash/s are not added to the network.

So if those 30Thash/s hit the network lets say November 7th 2012. We will see a new Estimated Difficulty the following week November 14th. Assuming the network speed at that time being 53.2Thash/s the difficulty would be 5,309,047.00

I expect that there will be another 30Thash/s getting onto the network during difficulty cycle that started November 14th.

By November 24th expect to see an Estimated difficulty increase between 8,302,870 and 8,703,651.

The max potential increase in dificulty cycle and assume that there will be 3 difficulty cyles within two 4 weeks from shipping would be 4x of the previous deficulty.

If we got a difficulty change in 5 days then 14 days the difficulty would be at a wapping 12,242,592 for the first difficulty increase (5-7 Days).
Followed by the next estimated difficulty 4x 12242592 = 48,970,368.


My personal numbers are calculated at the difficulty being as follows.

November 7-14th 2012 @ 5,309,047
November 18th - 24th 2012 @ 8,703,651

My personal maximum estimate because of other factors such as more ASIC's being added to the network then estimated.

November 7-14th 2012 @ 12,242,592
November 18th-24th 2012 @ 48,970,368

Note: This is my personal augmented reality perspective on the situation.

[squid]

I doubt we see any ASIC until Dec at the earliest, personally I think Q1 2013 regardless of manufacturer.