Topic: Total Hashrate Forecast Q1, Q2 2014 (Community work) - page 6. (Read 16695 times)

Prices will drop that low and much lower. Lets take BFL monarch as a baseline for a second. I've  linked this table before, that shows the BOM of highend video cards:

source: Mercury Research

If you eliminate the components you dont need on a miner (mostly GPU and VRAM) you end up with $47 for a 6970. A card which has comparable power draw and size as a Monarch (but which also includes a ton or irrelevant components and connectors like crossfire, DVI, HDMI, audio, PCIe 16x, ..  and a far more complicated 14+ layer PCB to accommodate the ultra high clockspeeds and bandwidth of the VRAM).

Which leaves one to estimate the cost of the ASICs. With a per 28nm wafer price of $4000 (50% above high volume prices), a 400GH hashfast golden nonce chip would cost somewhere on the order of $30 after packaging and yield. I assume BFL and most other vendors will be in the same ballpark, order of magnitude of $100 per TH for the chips.

Combine both, and add some minimum operating margin and you will see that $1000 per TH provides these vendors with gross margins that would make intel and nvidia green with envy. In high volumes, a 600GH Monarch probably costs BFL barely over $100 in marginal production cost.

At today's BTC prices you won't see anything getting turned off until we are in the 75 Ph/s region. Even then, the gear that is no longer cost effective to operate will be a small part of the total rate, so the result won't be noticeable.

A larger factor in network growth will be the time to break even on the hardware investment.  Puppet is very optimistic about price reductions, with his lowest hardware cost being less than 1/2 of the cheapest pre-order prices.  On top of that, deployments at any scale have substantial facility costs for racks, network gear, cooling and power.  These can easily approach $1000 / Th/s.

Without more massive bitcoin appreciation, the risk reward for buying hardware will look pretty ugly by the end of Q1 2014.  Paying 10s of thousands of dollars, 4 months in advance is tempting when you earn your money back in 2 weeks.  When it will take 6 months of hashing with stable difficulty and exchange rates, the smart money is going to step back.

Updated

Thank you everybody for your input. I will consolidate the data today and make a first draft. We can continue tweaking the numbers as time progresses.

Ya, but it seems clear we're not talking about a 20 ph network, but like 45 ph/s minimum just by April!?

Buy BTC in tripple digits while you can! Backing up the truck. Miners! Really?! Just can't help yourselves. And thanks for raising the value of my BTC!

Only that part which is no longer operationally profitable, ie costs more in electricity than it generates in mining. Very few, if any, asics would become operationally unprofitable in a 20PH network, unless your electricity costs are  high, or BTC value tumbles. Maybe some obsolete first gen Asicminer/avalon gear, which by then would form an insignificant percentage of the network. En even then, rather than being shut down, I suspect most owners will either continue operating at a loss or sell their gear to someone with cheaper electricity (or poor math skills).

Something to consider, if the total hash is 10 phs, and it doubles, effectively doubling difficulty, what % of the first 10 phs gets turned off?  Once the old equipment breaks even, it's house money, and I suspect only breakage or electricity>btc will be the only reason to decommission.  So does it all stay on for 2014? Or does much of it become too expensive?

This is pretty insignificant. Preamble is misspelled on the OP.  It just bugs me.

Its possible, though if I were to guess, Id say they are honest. BA is not a large company, they were the last vendor to announce their (28nm)  product, and as result had a tough time, certainly  initially, getting a meaningful amount of orders. I always  suspected they were one of the few companies with a realistic delivery schedule, but most miners got fooled by BFL, HF & Co who promised them delivery so much sooner.

ANyway, now that HF has fallen from grace, CT sold out for months, I do get the impression BA sales have picked up substantially recently, so maybe they will aim a bit higher for subsequent batches. Whether that will make them a player that needs to be counted with in the overall picture, I dunno.

800TH, that was much much lower than I expected. Could it be a case of understating the numbers?

About BA:
"We plan to manufacture 8000 ASICs in the first batch (800 Thash)."
http://ecointalk.net/topic/11-prepaying-full-or-50/?p=17

Datacenter availability, that's an interesting tought.
I agree with your post and we don't have any solid data so this is all going to be guesses of some sort.
22PH for BFL in Q2 could be a good guess (on the safe side I hope).

Do anyone have any ideas about Black Arrow? I read somewhere their first batch was going to be 8000 units, sounded like a lot of units/preorders to me at least if that meant the number for X-3 since I think they are now on batch 2. Anyway that's 16PH in Q1 if all is going according to plan. And maybe another 16PH for Q2 for batch 2. But I have no source for this.

I agree we wont reach the "end game" equilibrium in 6 months, it will take much longer to get close to it, but when you see how far we are from that equilibrium, then you must realize that only brake on network growth is manufacturing capability. And its hard to see just how limited production capacity will be 6 months from here, taking in to account fabs have virtually unlimited production capacity and there are plenty of high volume assembly houses that are starving for business with the decline in PC sales. Perhaps the most realistic limiting factor will be datacenter availability.


Yes, but no one knows what that will be. First time I ran the above numbers, I assumed a bitcoin price of $200. Its now almost 4x that, and as a result, you pretty much get 4x higher numbers. If you think bitcoin will go further up, just multiply (or divide) the numbers in the chart accordingly.


Problem is that we have no solid data for that and no source that doesnt have an incentive to understate the numbers.


My brother read an article in a semicon magazine about BFL. Unfortunately I cant find it. He said the article mentioned BFL had (pre)sold $180M worth of bitcoin asics. That was 2 months or so ago. BFL sold their 65nm gear for around $50 per GH. That would work out to 9PH. Thats roughly twice the hashrate I believe they could be responsible for today. Which leads me to conclude about half of that sales number is for yet to be delivered 28nm gear. So around $90M.  At $7.8/GH thats 11 PH. Lots of assumptions, but its a ballpark number, at least for their order books 2 months ago, before the bitcoin boom. Id feel comfortable doubling that number today.

Yes 20PH next July is smoking some weird stuff but so is 2000PH i guess.

I don't think the end game calculation will be a good prediction for the next 6 months. But it is interesting. But I guess you need to include the price end game for bitcoin also in that calculation. I also believe that the end game of mining will be mining companies with employees wanting pay, stockholder wanting dividents, replacement cost of equipment etc... so I electricity won't be the only cost it must cover, but it will of course be the largest one. 

I believe that in a 6 months timeframe estimation of production capacity and the current orderbooks will propably give the best estimate. I could be wrong of course.

But back to BFL is there any information, excluding from Josh who I think might over promise a bit, of how large their orderbook is and how large their production capacity will be when/if they start delivering. Is there any past estimation before the conversion to Monarchs?

BFL are infamous for overpromising, Im not going to bet on when they will deliver, I would never order from them, but the reality is also that its now a 100+ employee company (Knc is what, 10 employees?) that has been raking in preorders for an eternity now and have been shipping close to 1000 miners per day recently. Miners that are more complex to manufacture and assemble than Monarch. Considering they have not yet dropped their price one penny, I suspect they are not short on orders.

As for the combined network hashrate, IMO a better way to estimate it is simply looking at the profitability. As long as mining is substantially more profitable than it costs in electricity, the hashrate will go up as fast as all vendors combined can deliver. At todays BTC price, and assuming ~$0.1-0.2/KWH that means a race towards 500- 1000 PH. Anyone who thinks we will be at 20PH next July is smoking some weird stuff.

Here is a chart:

Puppet of course these numbers will be off but I think it would be worth the effort just to get the right order of magnitude. I would say the different views people posting here have of total hashrate in july differs from 20PH to 2000PH. We must be able to close that range a little bit.
Even if the conclusion is something like 100PH to 300PH (could be something different) I think it would have been worth the effort.

Cointerra is to low. I remember reading somewhere that they have batches of 1000, could it just be that darcimer's posting just is the number of units not the hashrate, meaning we double the numbers.

Regarding BFL, are volumes of >10x (18000 units) really realistic? This is question not a statement. Even if Josh thinks he can produce and sell 1000 monarcs every day, there must be a more realistic number for us to close the range between 20PH and 2000PH i july. One assumption could be that BFL will only be able to push on half hashing power of KnCminer before Q2 and set BFL to something like 7500PH TH in Q2. I might be completly off here. Discuss.

Many of your numbers are way too low.
First of all, AFAIK cointerra is making batches of >2PH each. At least their first, december batch was 2PH and I would expect subsequent batches to be larger, not smaller.

600 Monarchs per month is laughably low. BFL being BFL, its anyone's guess when they will ship, but when they do, their volumes are likely  >10x that.  Josh is on record saying they expect to be able to assemble over 1000 Monarchs per day. If their order books only had 2 or 3 days worth of production, do you really think they wouldnt drop prices? Just from undelivered 65nm orders that were converted in to Monarchs or cloudhosting I would expect several thousand units at the very least.

Anyway, while I appreciate the effort, the result isnt going to be worth the effort. These companies arent telling us how much they will produce, because those numbers hurt their sales. Expect them to understate their capacity if they even say anything at all.

@cloudhashing preparing to open mining facility near Dallas, with over $3 million in machines produced @cointerra
http://tinyurl.com/l7pybor

I tweeted this recently after reading an article by @zerohedge

3million purchase from Cointerra by Emmanuel Abiodun of CloudHashing & the Icelandic BTC Compound accounts for 500 machines... Not sure which batch. My guess is early.

I'm so glad this thread was moved and continued.
Thanks de_ixie, great move.

You can use this or not use it, none of this is new info, and I can't provide proof for any of these figures, except for the fact that the information is claimed to be derived from D&T's thread https://bitcointalksearch.org/topic/estimate-of-asic-pre-orders-13-to-15-phs-diff-18b-to-21b-by-end-of-2013-283820  and used in this spreadsheet here
https://docs.google.com/spreadsheet/ccc?key=0AqGoWjATRQK6dE1yTHpydkIwajhhUGF4ZGxFV2FDSVE&usp=sharing#gid=0
Which I can not validate except for cross referencing the two links.  I was hoping it might be a good addition to the starting point.

Thash:         2880 Thash
Origin:              Hashfast
Product:              BabyJets
Expected Delivery:   Nov and Dec batches (1000 each plus 880 MPP or reserve)
Scenario:              Maybe just late
Source:              Above- Same source for all- supposedly from https://bitcointalksearch.org/topic/estimate-of-asic-pre-orders-13-to-15-phs-diff-18b-to-21b-by-end-of-2013-283820 but could not verify all
Reasoning: Very late already, refunds due if they don't meet Dec 31 shipping, speculation that HF BJ's are already on the network at Eligius

Thash:         5500 Thash
Origin:              Cointerra
Product:              Terraminers
Expected Delivery:   Dec, Jan, Feb, March, April  batches(1000, 1500, 1000, 1000,1000)
Scenario:              Maybe late too, that's alot of work
Source:              Above- Same source for all- supposedly from https://bitcointalksearch.org/topic/estimate-of-asic-pre-orders-13-to-15-phs-diff-18b-to-21b-by-end-of-2013-283820 but could not verify all
Reasoning: rumblings about chip delays

Thash:         1800 Thash
Origin:              BFL
Product:              Monarch
Expected Delivery:   Feb, Feb, March (600 each)
Scenario:              Maybe late- Reputation for being late
Source:              Above- Same source for all- supposedly from https://bitcointalksearch.org/topic/estimate-of-asic-pre-orders-13-to-15-phs-diff-18b-to-21b-by-end-of-2013-283820 but could not verify all
Reasoning:

Thash:         4000 Thash
Origin:              Bitmine
Product:              Coincraft Desk and Rig
Expected Delivery:   March 2 batches(2000 each, possibly 3900 )
Scenario:              Realistic, currently shipping Antminers
Source:              Above- Same source for all-  https://bitcointalksearch.org/topic/estimate-of-asic-pre-orders-13-to-15-phs-diff-18b-to-21b-by-end-of-2013-283820 but could not verify all
Reasoning:

Thash:         1000 Thash- a guess? Needs more info
Origin:              Blackarrow
Product:              Prospero x-1 and x-3, also minion rental, cloudhashing
Expected Delivery:   Feb, March, maybe April too- claim to be all batch 1
Scenario:              No confirmation of how many units they pre-sold or plan to ship
Source:              Above- Same source for all-  https://bitcointalksearch.org/topic/estimate-of-asic-pre-orders-13-to-15-phs-diff-18b-to-21b-by-end-of-2013-283820 but could not verify all
Reasoning: Can't find any supporting evidence to this one, but have an active reseller here, and appear to be still taking orders, so 1000 sounds kinda low. Maybe they can give us a hint.