Topic: Guesstimate thread for total ASIC pre-order hashing power. (Read 7902 times)

Bcp19, you should be ashamed of yourself.  I would remove your posts but D&T has already locked and started a self-moderated thread, leaving your posts makes the necessity of it clear.

Locking.

https://bitcointalksearch.org/topic/estimate-of-asic-pre-orders-13-to-15-phs-diff-18b-to-21b-by-end-of-2013-283820

Then I suggest you do just that... *I* am the only person you have warned even though some others have done pretty much the same.  The difference is THEY are anti-BFL and that is all that matters.  I don't even HAVE a BFL hat on at the moment, but you have pre-judged me and are warning only me.

Take it off thread bcp and read the rules in the OP.  If necessary I will lock and recreate as a self moderated thread.   

You talk as though no other company would have done this.  If you simply open your eyes, you'll see that most of the companies operating at this time are doing the same thing.  Network hash rate is poised to skyrocket... while I will be shouted down since everyone thinks I am a shill and too ignorant to think for myself, I have alarm bells going off over the companies that claim they will guarantee a 90 day RoI.  If the network jumps by a factor of 4-8 over the next 2 months like a lot of people are predicting, that means those companies will have to then increase their product hash rate by that same 4-8 times which will then turn around an put an even bigger strain on the network.  Personally, if *I* wanted to scam people, I'd dangle that 'guaranteed RoI' in people's faces and then vanish in the night a month later.

So, shout me down, call me a shill and a sock-puppet and an ignorant fool... time WILL tell.

I’m wondering if Avalon's contributions are a bit conservative here. In a few articles from TGB, they consider the chips capable of OC’ing to 400 up from the nominal rating of 282 Mh/s.


It may be in error to expect every chip OC’d, but depending upon your intended use with these figures, you may want to account for some potential of higher rates out of Avalon.

http://thegenesisblock.com/latest-shipment-of-avalon-asics-could-increase-network-hashrate-by-500/
http://thegenesisblock.com/800000-avalon-asic-bitcoin-mining-chips-expected-over-next-month/

Also, just wanted to say thank you to D&T and the contributors here for this thread and data. This is an extremely valuable roundup of information that I wish I had a month ago. As this is relevant to every miner and prospective miner, I’d personally like to see this become a sticky thread. I’ve also used some of the data here in a difficulty projector / profit calculator google doc I just began to share.

https://bitcointalksearch.org/topic/m.3033042

I'm very skeptical of the claim that BFL produced 75% of the network increase in the last month. While ASICMiner's share of the network has dropped, I don't think it has dropped as much as it would have had they not been bringing a substantial amount of hashing power online. 100TH has also been coming online for the last month or so, and I think between the two of them they probably make up the bulk of the hash rate increase. It's a good idea to overshoot these estimates for the sake of calculating profitability, but I suspect that the 6-8x multiplier used for the BFL calculations is probably more like a 3-4x.

In either case, it is very clear that BFL has oversold their units well beyond the point of their own customers being profitable even if they were the only ASIC company out there, which is a very shady and nasty thing to do. I just don't see how they will last another generation of ASIC production given this track record. Many have talked about BFL being a scam of one form or another; while I just think it'd be far too elaborate a scam to actually fab chips and put out these products only to cut and run cashing out on their 28nm vaporware, watching them go under before shipping all of their preorders from Gen 1 seems like a real possibility. The effect for customers who lose their preorder money is the same of course...

Avalon dropped the ball with chip supply, BFL are in a position to capitalize on that if they wanted to.

There are already some BFL board projects just waiting on chips to arrive, which will add a reasonable amount to the net hash. I am sure lot's more DIY group buys would form if BFL sold 65nm chips at a competitive price.

I would expect that BFL have a massive markup on their 65nm chips, there is no reason why they couldn't keep selling them at a reduced price, whilst they focus on selling their 28nm product at a premium.

That would be an unknown in the TH/s calculations.

If any of the others are designing their own chips please provide a link and I will update the OP.  I am not familiar with any of them except BTCGARDEN and don't really know the details of that one.

yes I haven't taken into account orders that have been converted from SC to Monarch - although I doubt it's too much, and most of them I expect are post-April orders.

Will

If a mining op is buying rigs from a manufacturer I don't include it in the total (as it is already implied in the ASIC makers total). 

I did it for IceDrill because they have a special deal to buy chips not boards from HF, although in hindsight this probably just adds confusion so I will simply make a note of their purchase and increase HF total.

Will.  Sounds logical enough for me.  With the added note by Syke I am comfortable with 3,000 Thash.

At this point I can't see BFL ordering more 65nm chips given how aggressive 28nm competitors have become.  So I think it is likely they will eventually use all ~3TH/s of 65nm chips.  The fact that they are still advertising 65nm product probably means they are somewhat below that.  Even if people cancel I think all 3 TH/s will be used.  The chips still have value and once the backlog is either shipped or upgreaded if BFL offered 65nm devices in stock w/ next day shipping they could sell them.

This nice thing is this also helps with the Monarch math.  How much 65nm hashing power will the upgrades to Monarch remove?   I think it is plausible now that it is none.  Ultimately all that 65nm tech will be used, even if BFL needs to self mine.

Thanks to both of you.

A while ago they had 75 wafers on order, with 1000 chips per wafer that's around 300 TH/s. Unfortunately they haven't announced ordering any more wafers. They're going to have to order a lot more wafers to reach 4 PH/s.

it's basically a lot of assumptions, which is why it's more of a guesstimate than anything accurate.  Not sure showing my working will help other than to stress how much we don't know about BFL order queue.

I'm assuming that BFL have been 75% of the hashrate delivered in a three month period between June and now.  This is around 100 - 550 which is 450Th and 75% of this is around 330Th.

I'm then looking at the public order list for these months and adding up the Th based on MR orders until the current date for MR (28 July)- 22 orders and only 2/3 minirigs delivered (from Jody's blog they are delivering 500Gh at a time) so 15 * 1.5 = 22.5Th then minirigs form average of 277/(277+36+150) = just over half (60%) the BFL hashrate is minirigs so that's around 40Th (including others) delivered from the public list.

so 40/330 means that the factor of 'hidden' orders is around 6-8 times (depending on the estimate of 75% above).

so now take the total number of orders which is around 500Th and multiply by 6-8 and you get the 3-4Ph that BFL will deliver.

I know it's a TON of guesswork here - which is why I give an error margin of 3-4Ph from BFL.

Will

That is their stated goal and as a shareholder I agree it is necessary, but I am also concerned about the ambitiousness of it.  The recent rise in hashing power above 50 TH/s is a good sign.  If they can get the network stable above 50 TH/s for a couple weeks that would be a good sign they have the power and space to handle the upgrade.  Also I think efficiency gain will be lower than 16x.  If it is say 12x then it becomes even more ambitious.    Still the purpose of the thread is to document planned hashpower increases so I have AsicMiner down for 1,000 TH/s in 2013.

If you can show some links or references I will raise the estimate.  BFL isn't the only one who has been delivering hashpower.  I hated in school when teacher said "show your work" (because I could do it faster in my head), but "show your work".

BFL is likely to be more than 2PH.

The basis for this is taking the amount of hashrate that's been delivered to the network since around June time (when BFL started shipping a lot) and comparing that to the public order list - this gives a 'hidden' order list of around 6x - meaning BFL would be shipping between 3 and 4 Ph.

This estimate doesn't include any 28nm tech

EDIT: FWIW my mining investment decisions are based on a guess that there will be around 16Ph on the network by May 2014.

Will

ASICminer are going to have to get up to 1PH/s by Xmas to remain in the game.

Assuming they can get a 16x power reduction from 110nm to 28nm die reduction, they are going to need a 20x hash rate increase to stay where they are now. That would mean replacing their entire farm with 28nm tech and expanding it by Xmas.

just want to say thanks for the effort that everybody involved in the thread have made to gather this precious information. kudos.

Bitmine can be ignored for these calculations through January because they are already factored into Avalon and Bitfury.

Added thanks.

Updated thanks.

Is that y-axis difficulty or hashrate? I think that 10 Petahash/s is a solid estimate for the beginning of January, if it's difficulty you might be overshooting some.

Edit: looking at the value for 8/24 makes it clear to me that you're measuring hashrate.

http://www.coindesk.com/bitmine-to-drop-4phs-of-asic-power-onto-bitcoin-network-before-april/

Finally found a source for 1000th for asicminer, last paragraph
http://thegenesisblock.com/cointerra-expects-to-deliver-2-phs-of-asics-in-december/

Anyone have a good idea on how to figure out ASICMiner sales other than try to look historically for every group buy and direct sale in the past couple months?

If not posting links and totals form various group buys would be a start.  If any reseller wants to help me out that would be awesome.

Something like

Thanks, that is a new one.  I think I will change the OP to break it into deployed in 2013 and deployed in 2014.  The reason is that due to falling revenue per GH/s each month it is more important to get the short term accurate then the long term.  4 PH/s in March really doesn't affect someone who has a current pre-order because honestly if you haven't hit 100% ROI by March you probably aren't going to and if you are profitable you likely will make most (75%, 80%, 90%) of your lifetime revenue.

http://s21.postimg.org/mqfwmlz93/Photo_2013_08_27_4_50_54_PM.png

I edited a little bit, as per discussion. Don't know if it is more proper to edit post or make new post, also,  might have undersized.

Feel free to suggest changes.

I love the chart, it is very well made.  I do have to admit that you predicted the 'explosion' to happen about a month after I did, but we have more information at this time than I did when I made mine.  Very well done.

Bitmine?

http://www.coindesk.com/bitmine-to-drop-4phs-of-asic-power-onto-bitcoin-network-before-april/

Maybe late, but plenty.

It was my first draft, just wanted to get some numbers on there just to help people visualize.

I will edit some this afternoon.

I have KNC down for a total of 2 ph, it probably looks like more because when there is an addition underneath it throws everything else above it off visually. I picked 2ph instead of 500 th just to go for worst case scenario.

Hashfast and icedrill, used their numbers, will add more for January hashfast. Bitfury, pretty much just used this thread's numbers.

Why do I think Avalon will have so many chips? Their next gen coming out in October, no preorders, lots of nre money from previous success and mining with people's equipment or not shipping to keep difficulty down. I will adjust downward.

Asicminer, I am fairly confident in that number. Friedcat knows what he is doing and has more money than anyone. Comitted to 20% of the network. Shrewd business man. 28nm chip 8x more power efficient, so they can be up to 400th right there. Double that with their leasing plan, and throw in 200th for sales.

As for shipping dates, purely just guessing. It would take a lot of research to nail down better estimates, which most people would disagree with anyway.

I will pretty it up though and post update.

Interesting chart.  IMHO (which doesn't really mean anything don't take it the wrong way) I think it is too aggressive.  I would consider it a worst case scenario.  I find it unlikely the growth would be more aggressive than that.  I added it to the OP.  If you can provide numbers I would like to add that as an alternate viewpoint for the OP.

Some comments in no particular order

KNC: Shows shipping 2.2PH/s in roughly 30 days.  Based on breakdown on known orders, the average unit is ~0.25 TH/s.  That would mean shipping ~8,880 units in 30 days.  While not impossible it seems improbable if for nothing other than a manpower perspective.  To go from zero to ~300 units per day would be impressive.  Also I am interested to know where the 2.2 PH/s.  Based on their probable die size that would be a massive number of wafers and represents about $38 million in sales.  For the record I think other estimates of very small die are inaccurate.  Given the package size and power consumption it would seem to be unlikely.

Avalon:
~1.2 TH/s by end of year is about 3x all known rigs and chips.  Curious why you would think Avalon will sell that many chips?  Especially now given their massive delays, lack of price drop, and poor "rep" in the community.  Since Yifu update, among the major bulk orders a grand total of 3 orders totaling about 24K chips have been shipped.  That works out to about 3.3 TH/week.  To ship another 1000 TH by the end of the year (~18 weeks) would require shipping 20x as fast.  At the current rate it will take two years just to ship the known 275 TH/s of chip orders.  I am not sure what the problem is with Avalon but I don't think they will sell that much capacity nor will they ship it that fast.

ASICMiner:
1TH/s by end of year seems ambitious given they have had difficulty keeping the 50 TH/s existing farm operating anywhere near peak capacity (falling from 47TH/s to 30 TH/s over the course of two weeks and despite recovering in the last week not exceeding 40 TH/s).  BTW I am an ASICMiner shareholder and I believe in Friedcat but I would think 1 TH/s by end of year would be a best case scenario for shareholders (worst case scenario for everyone else) and I wonder how likely that is.  Granted ASICMiner additional capacity is likely a 2nd gen chip and thus should have on the order of 4x (55nm) to 8x (28nm) power efficiency and density (hashpower per square foot of warehouse space).  Still 1TH/s even w/ 2nd gen chips would represent growing the farm's footprint (space & power) 500% to 1000%.  Given the difficulty in maintaining 50 TH/s today I am not confident that will be easy.  A lot could change in the next couple weeks though.

Bitfury & Hashfast:
The growth seems probable but the max capacity seems light (kinda hard to tell with the scale of graph).  For HF they will have to have another 880 TH/s to cover excess chips for MPP.  Either way those chips will be used in January.  Either they 1st batch will make ROI and HF will sell the excess chips or they won't and the excess chips will be given out as part of the MPP.

Cointerra:
Could be the graph but it looks like more than 2 PH/s.  I also think showing delivery starting 1 DEC is at this point impossible; "in Dec" is ASIC code for 31 DEC (and likely 1-2 weeks late but that will be close enough that we don't look that bad).  Similar to KNC I don't think 2 PH/s per month in delivery is likely.  Although based on KNC "rollout" we will have a better idea for Cointerra. 

Still I think it is great, it got me thinking about other options.  It also points a pretty gloomy picture as a worst case scenario.  If it unfolds like that I don't see anyone except the chip producers having a % ROI.  Any chance you could share the raw numbers behind the graph and your rationale?  I am interested more in the thought behind the numbers than the actual numbers.

Here is my guesstimate:
http://s11.postimg.org/awaew7fpf/9_CYQFOR77_tztv_Jrpiy65_Pz_U1q09_L2_Siv_G37g0_Ol_OD4.png

No announcement of tapeout, but there was this thread on Reddit: http://www.reddit.com/r/Bitcoin/comments/1l0plk/an_insiders_take_on_cointerra_the_bitcoin_mining/

It looks like they have VC funding, so its possible that they will get the chips made without preorders, but I agree that they will probably not make what they think off each mining rig. Curious if you think VC funding will make this a more sure production?

This is where it start to get confusing.  Can cointerra produce 2 PH/s without pre-orders.  If not then do they have enough pre-orders to start production.  Given the almost universal negative response I would think not but then it starts to get real subjective.  I guess I will just mark em down for 2 PH/s.

Thanks.

On edit: maybe should indicate if tapeout has been completed.  That is the final large fixed cost.  A company which doesn't complete the tapeout may never produce an ASIC but once taped out it is highly unlikely a company won't be able to sell all of its initial lot.  Maybe not at the price it wants, maybe the company even loses money on the deal but ultimately that hashpower is going to make it to the market.

If someone wants to rise the table to show "status" and anticipated delivery I wouldn't complain. If not I will do it this weekend.

Well it looks like Cointerra is announcing their total hashing power... http://www.coindesk.com/cointerra-cuts-price-of-terraminer-iv-bitcoin-mining-rig/

Looks like 2P/H total from them, with a price drop. Still not looking very profitable, but maybe things will change.

1000TH by ASICMiner
https://bitcointalksearch.org/topic/m.2610794

+Alydian http://www.coindesk.com/alydian-targets-big-ticket-miners-with-terahash-hosting/
+labcoin
+cointerra
+xcrowd

I think you are pretty close on your estimate of total starter kits sold (globally), but we sold far less full kits.  Figure about 50-60 full kits for August delivery worldwide.

In the long run yes however I was more looking for the short term.   Once shipping in volume margins on hardware is going to collapse and trend towards a small markup over cost.  Still even if raw silicon is $0.20 per GH/s the overall system is going to be higher.  Even simple stuff like high quality DC to DC PSU aren't cheap and can easily be more than the ASIC itself.  Throw in AC power supply, case, cooling, testing, assembly, labor, etc and you will be hard pressed to build a system below $1 per GH/s.

Vendors who are past their NRE cost can provide practically unlimited quantities of chips at the cost that is orders of magnitude below today's prices. Thus, the answer to your inquiry is mostly in the domain of psychology of the miners, not in the domain of engineering of the vendors.
The hashing power that will be coming online through the gaping anus of all these preorders is elastic in the sense that vendors will simply keep selling as long as (pre)orders are coming in. Once the supply of suckers starts diminishing, the vendors will simply drop the $s/hash ratio. This will attract new customers, and also force some of the old ones to keep ordering so they don't drop out of the game (monarch anyone?). Then the price drops again. And again. And again. Only when we get close to the production cost, and vendors start operating on thin margins, will the game of power efficiency begin.

What does it cost to produce high volumes of ASICs? A couple of tens of dollars per chip, which is thermally limited to a couple of hundreds of watts, which currently might provide a couple of hundreds of GHash/s. There you go. Let's say ~0.2$s/Ghash. Next, we account for the RoI period miners are comfortable with - a year maybe? That's 1.3M coins up for grabs. Their value is anyone's guess, but let's say $400 per coin in the next year or two. That's half a billion dollars to be made, ignoring fees. The miner's investment then translates into 2E9 Ghash/s. Give or take. Once we hit that, the J/hash will become more important, and that's a whole another amazing story full of exotic technologies, exotic places, and politics.

Added.  Thanks.

Still waiting.

For posterity.

My projections have us hitting an average diff of 411,484,534 in December 2013. This would be an average network speed of 2.94Ph/s and would mean that most ASIC equipment out would be unprofitable to mine with shortly afterwards (2 to 3 months) unless there is a price spike or electricity and hosting get a great deal cheaper.

I can't find a source to back up 1000th for asicminer by the end of  the year, I am sure I read it somewhere, but it must not have been official. A lot of people will have alotmore to say than me, but a few points:

Friedcat seems pretty motivated to keep a roughly 20% share of the hashrate. While they might have dropped to 10% lately, I am pretty confident they are ready for what's to come.

Friedcat would have access to more money than any of the other manufacturers.

The next generation of their hardware is known to be in the works.

They plan on leasing out their hardware, which I would assume means they will have a lot of it.

And one last bit of evidence would be their share price. If everyone lost confidence in friedcat's ability to deliver, their shares would be worthless. They have dropped recently but only because of their recent lull.

—> 1400 * 8 * 2.9Gh/s = 32.48Th/s

Meaning, don't forget the Russian BitFury projects.

1) Has anyone compiled a guestimate of the amount of pre-orders.
2) I did a breakdown in a previous thread I cannot find ATM.  Worked out to an average .1TH per BFL order and ~70k orders using the ~4900 known orders.
3) Refunds were not taken into account due to lack of information.
4) Monarchs were considered the "and rising"

Curiously, if you want to choose a single value to estimate beginning of January hash power, 50% per month is not nearly enough  

Assuming 50% till January:

500*1.50^4. =  2531.25

To get a value close to the sum (assuming "only" 500 TH/s for KNC) to be delivered till January you need about 68%:

500*1.68^4 = 3982.97

But choosing a single value is not the purpose of the topic, sorry.

Nice analysis. Much better than the "let's add 50% every month until forever".

Was talking just about the KNC numbers.

How do you get 2000 as the limit, when the running total is 3900, or did i miss something

Wow! Sweet analysis DeathAndTaxes. Those extra considerations do constrain the die size quite a bit more to the upside. 2 PH/s is definitely an upper limit likely to be much above the shipped total (nevermind 3). Learning quite a bit here.  



  

DT, thank you for this post...  your one of the few 'givers' here.

at 3900 Th, that puts diff at 540M

So, can we say that if everyone delivers, we will be at 540M near jan 30 2014?

or if 70% of the estimate gets delivered we will be around 400M diff near the first of the year?

edit:  I personally have been doing my guesstimates for profit starting Jan 1 at 2500 Th, adding about 300Th week.  Looks like i will have to bump my starting point on jan 1 to 3000Th

No.  Will add.

If their die size is indeed that small then 2 PH/s seems likely but I don't think the die will turn out to be that small.  I will make a note in the OP and let others weigh in though.  Given the use of 55mmx55mm package and the power consumption I believe the die is much larger at least 20mm x 20mm and that is being conservative.  That would put it at more like 15 TH/s per wafer or ~375 TH/s per batch.  Two runs would be up to 750 TH/s although my guess is the die is larger than 20mm x 20mm.

The package:
They are using a 55mm by 55mm package.  With flip chip packaging it is technically to fit an 11mm die to a 55mm package but would be rather bizarre. Yes using a larger package helps heat dispersion but it has diminishing returns otherwise CPU and GPU would have giant packages the same size as the heatsinks they ultimately will need.  

The power:
The quote estimates size by using BFL logic and shrinking one core from 65nm to 28nm and then scaling out.  That assumes KNC will have comparable GH/mm2 as a 28nm BFL but if true they should also have comparable GH/W.  However BFL is 6W/GH (at 65nm) and scaled down would probably <1W/GH, yet KNC is reporting 2.5 GW/GH.  There isn't a 1:1 relationship but power consumption generally scales with die size.  Another data point would be HF which has a 18mmx18mm die (28nm) which is ~1/3rd the quoted size yet has a power consumption of 0.6 GH/W.  To believe their chip will be 1/3rd that of HF but use 4x the power seems less improbable.

So would you say the above is too crazy?

Did this wafer reasoning apply to other manufacturers?

Maybe they ordered only one batch of 25 wafers which would be only 1.5 PH/s and thus only 3 times your estimate. Lets say there's a lot of inefficiencies and put it at 1PH/s? 50 wafers plus inefficiencies (66%) ~ 2 PH/s ?  Been looking forward for a while to get someone to comment on the wafer estimate by HyperMega.


 

Did the bitfury numbers include the 100TH (now 200TH) mine project?

Good point I hadn't really considered any other mining bonds/companies because they are generally already included in miner/chip sales however this is an exemption.  I added the lower end of their projected range to the OP.

Didn't hashfast partner with icedrill for 250-363th for November? (https://bitcointalksearch.org/topic/bitfunder-icedrillasic-ipo-235-thash-mining-operation-powered-by-hashfast-269216)
Friedcat, he has more money than anyone, has said he would have a petahash ready for the end of the year. (Looking for source).

So basically, if you have a BFL(really F**Ked) or Avalon order(super F**Ked) or miner(about to be F**ked), you will be paying out of your pocket worse than someone with a Folding farm(or some other form of donated distributed computing) to mine within the next year. Looks peachy.

500 TH/s makes sense to me.  Bitfury capped their Aug sales @ roughly half that.  For startups there is limit on how many orders can be processed in a timely manner.  Even mundane things like assembly, testing, packing and shipping.  BFL with their ~2PH/s obviously doesn't care they won't be able to deliver all thoses orders for some time but buyers are more sophisticated now.  KNC indicated two days in they had sold out of units available in Sept and they haven't indicated that for Oct yes.  So at most two months of orders, 250 TH/s ea.  Yeah that works.

I think it is a good starting point.  I would want something more credible to move it higher.  It gets us within an order of magnitude.  It is highly unlikely they only sold 50 TH/s or 5 PH/s. 

I wish buyers would demand companies provide transparency on units sold (like HashFast did).  New buyers have a right to know how what slice of the pie they are getting, and existing buyers will want to new buyers to know this to dissuade them from buying.  Only the company benefits from hiding the information.

You should put down knc for some starting number and see where the discussion goes from there. 500th can't be an underestimate.

Got a link or breakdown.

BFL: 7PH and rising.

re: Bitfury here in NA

From what I remember Dave (MBP) sold, 300 August 400GH kits and 300x25GH August kits.  About 100 October 400GH kits and 150x25GH October cards and another 150 of H cards each 25GH.
Chips???  I remember rolls of ~3500 chips orders vanishing from the store like candies.  Maybe people on his thread remember the exact number of chips bought. There were some July delivery rolls for $200K+, but I don't remember how many were sold.  Ask Dave...

And then you have European distributor to add to the total.  My guesstimate is around 300TH being added from the initial Bitfury August sales and another 500TH+ for September/October deliveries (those are not realized but I suspect there will be rush to buy once August deliveries ship).

There was a comment somewhere about the factory assembling 500 H-boards a day. Nothing about start date and end date though.

lol?

12.5Th/s of BitFury being assembled per day

The latest spreadsheet out of Avalon says there are at least 970,000 chips ordered.
That is roughly ~250TH/s.

Bitfury has delivered at least 40TH/s but lord knows how many pre-orders they have taken.

Avalon Units (Batch 1, 2 & 3):  

Batch 1 & 2 :  63 Thash

Batch 3:    60 Thash

Avalon Chips:  At least 150 Thash    ~274 Thash

The idea
The purpose is to try to guestimate (and yes that means error) how much hashing power is in "the pipeline".  If there were no pre-orders every unit sold means a unit hashing and then difficulty rises a little and that makes new sales slightly less attractive so sales are slower.  This economic feedback model constrains hashing power to overall network efficiency, electrical cost, exchange rate, the risk premium miners are willing to accept and the time value of money.    There is a reason GPU mining never allowed the the network to reach petahash scale with miners, mining away at massive losses hoping to make it up with future exchange rates.  The problem with preorders is it breaks the feedback model and that is a great thing for hardware vendors; they can sell more hardware, earlier and at higher prices.  You can tell people difficulty will go up but there is a lot of uncertainty in how much and how fast.  That uncertainty favors the vendors.  Not until miners believe there is no profit and stop buying will prices fall.  Accurately projecting growth simply by looking at prior growth is probably an exercise in futility.

So lets look at it from another direction. If you know all pre-orders are X PH/s and you assume they (or most of them) will be delivered over some period of time (say between now and end of year) then you can come up with a more realistic curve for the next couple months.  You know hashrate now and you can project hashrate at the end of the year and then fill in some likely curves.  For that we need an idea of how much hashing power has been presold .  So throw me your cites, guesstimates, and official numbers.

The rules
1) To keep this thread from derailing please leave the "xyz is a scam" for another topic. On edit: I wish I had made this moderated.
2) A good starting point is the total pre-ordered amount.  This is likely unrealistic but we need a starting point.  Later that total can be discounted by the likelihood of fraud (the "coefficient of scamming").
3) If you have a reference or cite (even unofficial) to back up a guestimate please link to it.  If not a reasonable explanation is more useful then just posting a number
4) We can safely assume that all non-ASIC hashrate will go to zero so no need to break it out between delivered and ordered.  Eventually hashrate ~= total pre-orders.
5) No idea/number is bad.  Please be respectful.  Honestly nobody knows except the chip makers and most of them are keeping quiet.

The running total

Promised delivery by December 2013

Post 2013 rollouts






Relationship between difficulty and hashing power

Upper limits on difficulty based on hardware efficiency:
Miners are unlikely to mine when their electrical costs are higher than the value of BTC mined.  This limit can be called the electrical break even point and is based on:
a) the current exchange rate (USD per BTC)
b) the hardware efficiency (J/GH )
c) the miner's electrical rate (USD per kWh)

When hashrate/difficulty gets high enough it will cause the least efficiency miners to idle thus creating a sort of replacement cycle (i.e. x GH/s new efficiency hardware causes Y GH/s of older less efficient hardware to idle).  This should slow growth significantly because the returns on new hardware will be low, miners will be exposed to the bad news of less efficient miners being forced to idle and X GH/s doesn't mean the hashrate only rises by (X-Y)/GH.  It also illustrates the improbability of difficulty power growing exponentially over a long period of time like a year.  For example 65 million difficulty gaining 75% per month for a year results in 50 billion difficulty.  The electrical cost even at 1W/GH and $0.10 per kWh would >$250 per BTC.  

A related thread on the break even point is here:  Break even difficulty by hardware efficiency (power cost = value of BTC)



Alternate view
Here is an alternate visual representation by gkm22d.  I don't agree with all the capacities but I would consider it a worst case scenario (delusional miners may wish to close their eyes as this may be painful)




[1] https://docs.google.com/a/nacrypto.com/spreadsheet/ccc?key=0AiLYkKIHJaIsdHpIaGdUOWRYVUdncTNpNlVKbVhCbEE#gid=0  970,000 chips @ 282 MH nominal
[2] 550 orders @ 400 MH nominal
[3] Based on report that Dave (US distributor) sold out of their allocation of 300 full systems and 300 starter systems.  Oct is not sold out but conservatively it will if/when Aug deliveries are made.  I will assume that the EU distributor received an equal allocation (an assumption based on bitfury facing unknown demand and users in both markets).
[4] http://bitcoin.stackexchange.com/questions/8577/how-much-asic-power-has-been-or-is-being-shipped-in-2013  Crude assumption based on distribution of wait list (hashing power per order) and number of orders.  2PH/s is guestimated based on (avg GH/s per order of known orders)*(num order numbers)*(1/3 to account for unpaid/test orders).  Monarch is highly unlikely to ship in volume (if at all) in 2013 while upgrades cancel the existing 65nm order which would reduce the amount of 65nm pre-orders.
[5] Guestimate. https://bitcointalksearch.org/topic/m.2994436
[6] https://bitfunder.com/asset/IceDrill.ASIC
[7] http://www.dpcapital.net/blockchain/?hours=336
[8] http://en.wikipedia.org/wiki/Occam's_razor
[9] Reduced to 30 full systems in Aug for both US and EU distributors https://bitcointalksearch.org/topic/m.3010364
[10] http://www.coindesk.com/cointerra-cuts-price-of-terraminer-iv-bitcoin-mining-rig/ "In Dec" without a specific date can mean as late as 31 DEC.  Given that and the tight schedule and the fact that even a small delay would push it into 2014 I included it in the 2014 group.
[11] https://ghash.io/
[12]HashFast MPP will issue miners up to 4x their initial hashing power if 100% ROI is not acheived within 90 days.  Even if MPP is not needed, HashFast would need the chips in reserve and any chips not paid out in the MPP are likely to be deployed as quickly as possible.
[13] http://www.coindesk.com/bitmine-to-drop-4phs-of-asic-power-onto-bitcoin-network-before-april/
[14] http://thegenesisblock.com/cointerra-expects-to-deliver-2-phs-of-asics-in-december/ (Numerous references, Cointerra 2PH/s, KNC 0.5 to 2 PH/s, AsicMiner 1 PH/s, Bitfury 0.5 PH/s, Avalon 0.32 PH/s
[15] Upgraded BFL estimate from 2 PH/s to 3 PH/s https://bitcointalksearch.org/topic/m.3029092 (see also the next two posts)