ASICMINER has said that their next gen hardware total deployment will be somewhere between 2 to 20 PHps. So not much help there. I think 10-15 Phps is most likely.
In fact I've used 15 Phps from all major companies in H1 2014, in my calculations. If that's accurate (which it probably isn't), the difficulty should be about 10 billion by summer. After that I think the difficulty will plateau.
Topic: Total Hashrate Forecast Q1, Q2 2014 (Community work) - page 7. (Read 16716 times)
Great thread! I have been looking for this. Just using 30% diff change is not realistic in the long run and probable not in this timeframe either.
Let me start with some easy ones:
Thash: 4200 Thash
Origin: KnCMiner
Product: Neptune Batch I (customers only)
Expected Delivery: Until Q2
Scanario: Realistic
Source: https://www.kncminer.com/categories/miners
Reasoning: Probable going to be shipping early Q2 (since they expect to ship all three batches before end of Q2. They have sold 1200 with a low estimate of 3Thash so let's assume they run at 3.5Thash. 1200*3.5 = 4200.
Thash: 4200 Thash
Origin: KnCMiner
Product: Neptune Batch I (new customers)
Expected Delivery: Until Q2
Scanario: Realistic
Source: https://www.kncminer.com/categories/miners
Reasoning: Same reasoning as above.
Thash: 4200 Thash
Origin: KnCMiner
Product: Neptune Batch II
Expected Delivery: Until Q2
Scanario: Realistic
Source: https://www.kncminer.com/categories/miners
Reasoning: Same reasoning as above plus I assume that they will eventually be sold out.
Let me start with some easy ones:
Thash: 4200 Thash
Origin: KnCMiner
Product: Neptune Batch I (customers only)
Expected Delivery: Until Q2
Scanario: Realistic
Source: https://www.kncminer.com/categories/miners
Reasoning: Probable going to be shipping early Q2 (since they expect to ship all three batches before end of Q2. They have sold 1200 with a low estimate of 3Thash so let's assume they run at 3.5Thash. 1200*3.5 = 4200.
Thash: 4200 Thash
Origin: KnCMiner
Product: Neptune Batch I (new customers)
Expected Delivery: Until Q2
Scanario: Realistic
Source: https://www.kncminer.com/categories/miners
Reasoning: Same reasoning as above.
Thash: 4200 Thash
Origin: KnCMiner
Product: Neptune Batch II
Expected Delivery: Until Q2
Scanario: Realistic
Source: https://www.kncminer.com/categories/miners
Reasoning: Same reasoning as above plus I assume that they will eventually be sold out.
reserved
Preample
Basically, this thread is a continuation of DeathAndTexas's thread using the scheduled/expected delivery date of all main ASIC players to assemble a somewhat valid forecast of total network hashrate. Personally, I think this is the most appropiate way to do it and I highly appreciate his work. I will also copy the basic design of his thread - it is in good shape and people involved earlier will recognize it fast. I was hoping he would continue updating his thread on a regular basis - this said, I have no intention in stealing anyones idea - I am just interested in a valid hashrate forecast. Mr DeathAndTexas - in case I spit in your soup here - feel free to drop me a PM. Link to original thread -> Original thread
Goal
Produce a hashrate forecastQ1 2014 (31.03) and Q2 2014 (30.06) based on the delivery schedule of main ASIC companies (and other sources!?) for that timeframe by using community knowledge. Starting point is 01.01.2014 with a hashrate of 11 Ph.
Method
A) In case you have an indication/ wild speculation/ backed up source of how much Thash will be released by a specific entity post your estimate in the following format:
B) Any estimate is welcome. In case you want to contradict a statement please behave mannerful and explain why your estimate differs
C) Always try to include a source and a statement about your scenario assumptions 1) Lower tier, 2) Mid tier, 3) Upper tier
D) This thread is not to discuss any specific products in terms of ROI. Whether Miner xyz will break even is offtopic and will be deleted
E) Any post containing unrespectful behavior or other offtopic discussions will be deleted
F) Before posting check out the scam list first! -> Good effort!
G) I am open for any feedback or ideas how to improve our forecast approach. Don't post it here. Leave me a PM.
H) No one can predict the future. Let's try to generate a realistic case. In the end we will most probably have a range of xxx-xxx Thash.
Total Hashrate Forecast Q2 2014
Based on delivery by end of Q2 2014:
Conclusion, Chances & Sensitivities (Preliminary)
Rather than just posting a number I prefer to lead the discussion based on a range or tier system. For the first draft of this conclusion I introduce the lower, mid and upper tier.
Lower tier:
Upper tier:
Mid tier:
First draft - Feel free to discuss this openly with your insights! Still have the gut feeling our Mid tier is too high. I expected something around 7b diff earlier.
Version
- 27.12.2013; start of discussion
- 30.12.2013; first rough draft; removed Q1 FC; set focus to Q2 FC; renamed scenarios to "Lower, Mid, Upper"; draw first conclusion
- 07.01.2013; Hashfast failed to provide, but for now -until clarifiaction of situation - lets keep FC like it is; Some unknown source keeps adding massive hashpower; We have to observe that trend; Next review around end of January; By then, we will have a lot more information for Asicminer, Bitmine, Hashfast, Cointerra etc...
Next Step/ To dos until 07-09.02.2014
- Check batch definition of each vendor - I am not 100% sure whether I did mistakes (by: Community)
- Check & discuss lower/ mid/ upper figures (by: Community)
- Ongoing search for every tiny bit of information (from mails, forum, vendor hp, IRC, media) to verify our data (by: Community)
- Update tiers & conclusion including aboves findings (by: author)
Sources for forecast data
[1] https://www.kncminer.com/news/news-69
[2] https://www.kncminer.com/news/news-72
[3] https://www.kncminer.com/categories/miners
[4] https://www.kncminer.com/news/news-73 -> Batch size is pure speculation. Any ideas?
[5] Statement Friedcat in Asicminer thread on bitcoin.org -> Someone has the link?
[6] 550 Units@ 400-600 -> HF Order Chain information; Speed still not verified
[7] max. 400% of BJ speed-> Hashfast ToS
[8] No figure available; Is currently selling batch 4, which is guaranteed to ship by March 31, 2014; Once product is debugged additional batches will follow for sure
[9] http://www.coindesk.com/cointerra-cuts-price-of-terraminer-iv-bitcoin-mining-rig/ -> The 2 Ph is Dec and Jan batch???
[10] No figure availabe; following batch 1; CT is currently selling batch 5; Cointerra has without a doubt massive capacities; any ideas?
[11] http://www.coindesk.com/bitmine-to-drop-4phs-of-asic-power-onto-bitcoin-network-before-april/
[12] Pure speculation on basis of production capacity; I read they have capacity of around 50 rigs/ day, 20 working days, 2 TH biggest unit = 2 Ph/ month
[13] Pure speculation; No one knows right now. Josh claims capacity of 1000/ day, 20 working days, 600 Gh unit = 12 Ph/ month; From the past - Company has reputation of beeing late and overstating its capability. Nevertheless, they already deployed substantial hashing power & future potential should not be underestimated
[14] http://ecointalk.net/topic/11-prepaying-full-or-50/?p=17
[15] Any information??? Which batch is selling right now? How many units?
[16] Cumulation of smaller vendors; hidden Island operations; Start-ups whatever; safety margin
[17] Broken gear; numbers wild speculation
Let's do this.
--------------------------------------------------------------------------------------------------------------------
Update 11.02.2014
Finally I found some time to scramble the numbers. We got a lot of new information and developments. I try to cover them all. In case I miss something crucial feel free to give me the hint. In addition - before I start with this - at some point I will surely drop negative/ positive remarks about vendor x or y. This is not to attack/ promote the operation or the people behind it. This is simply the attempt to transfer their behaviour into some viable numbers. Fans of club A or B. Please do not feel judged - try to see this in the right light: We want a solid FC here!
Let's start. First - I want to check what happened between our starting point 01.01.2014 (11 PH) and the last adjustment on 05.02.2014 (18.7 PH). In this period roughly +8/9 PH got added to the network. Our initial guess of 10 PH per month up until July was surprisingly accurate for the first month. We all can agree that no major player in our spreadsheet shipped significant amounts. We saw some Hashfast/ Cointerra units in the wild but major consensus is that they have not started mass production yet.
Conclusion: All added network power originates from the position "Others" in our spreadsheet. Who are the players in here? We hear a few names. Most prominent should be Avalon chips (6 PHs sold), Bitfury, Antminer, 21e something... and maybe a few other hidden projects. Bad news is we have no deeper insights on the developments here. Good news is we dont need all the details to adjust our FC accordingly. It should be clear by now that my initial number of "Others" was WAY, WAY, WAAAAY to low. Given around 8 PH originates from "Others" in January I set FC to approx. 5-10 PH per month from now on only for this position. [1]
Let’s check step by step what our main spreadsheet positions are doing. From top to bottom.
KnC:
Will not deliver any additional Jupiters. Maybe a small batch of upgrade modules which will not have significant effects on the hashrate. They started to build an arctic mega-datacenter on the dark side of the Neptune to cover potential losses for their 20nm customers and move towards cloud based mining in the future. Nevertheless, they claim to have made significant progress on their development. I will only remove the Jupiter position in the spreadsheet and - for now - leave the "Neptune-Batch" position untouched. My feeling is, they either provide the physical Neptune in Q2 (unlikely) or the equivalent in virtual hashpower (more likely). Anyway, this does not impact our FC. I know there are people out there with different opinions about what KnC will do or not do. Still open for discussion here. [2]
Asicminer:
The quote of 2 to 20 PH for the initial batch of their new chip is still up to date BUT - they also announced to delay their tape-out to mid Feb. I am not an engineer - but my understanding here is: From tapeout to working chip around 60 days (mid April). Then they have to sell their chips, ship it, people need to start developing miners based on their chips and so on..... From observing their thread I also have the strong feeling their management lacks basic customer/ shareholder care, fast decision making and clear strategy. For me these are strong indicators for a delay. Although their chip looks very potent on paper - I just dont see any significant volume from AM in our observed timeframe. This said, AM FC is reduced significantly. [3]
Hashfast:
I guess not much to say here. They messed it up big time. As far as I followed the 1m+ pages in their tread they managed to ship their initial batch without MPP. Don't know whether they started batch 2. Clear is: They have engineering problems, they have court action going on from changing their ToS several times after sale and they suffered BIG reputation damage. Basically, they are out of business or very close. Only a brain dead monkey would buy from them in future. Maybe they recover, maybe not. Big potential to reduce FC. As of now - I will remove batch 5&6 from spreadsheet. Let's wait and see what they will do in Feb/ March to judge further. Maybe they mine with own chips? We simpy do not know. [4]
Cointerra:
These guys were said to have the biggest potential of the new vendors showing up. Solid staff, solid financials and overall profesional appearance. Fact is they are also late on their initial batch. As far as I followed their thread they are still struggeling to ship out their December units. We mainly see some guys shaking hands with the CEO for marketing purpose - no mass volume. Nevertheless, first units appear to be in hand by respected members of the community.
What else? Problems (challenges) with certification, problems with performance of the unit, problems with power consumption, problems with the boards. They have a working product: Yes. They do not have it shipped in masses and clearly under the expected performance. I still think these guys will pull it off and deliver what they promise soon. But I will reduce the FC due to the 20% less performance of the units. [5]
Bitmine:
Hmmm - I clearly have no idea. They are late yes. There seem to be some working chips yes. Do they seem to be profesional enough to ship in large quantities? - maximum chance 50% chance imho. Behaviour from the last 1-2 months: Strong lack of transparency and delaying tactics. They are buying time. Their customer base starts to get nervous. Let's see and wait. FC unchanged. [6]
BFL
Just dont want to comment on these guys. I will keep FC like it is. Next time maybe. Still see significant space for reduction here. The monarch looks like it will turn to molten lava in the second you power it....
blackarrow:
Another company basically predestined to fail. Horrible customer support. Last minute delay from February to MAI for their initial batch. Their chips/ product will be worthless by then. Customers start to jump the sinking ship. Chance of going bust within 6 month: Very high. Just my judgement. Anyway, the shipping volume we have in our spreadsheet is very low. I will let it unchanged this time. Next time we may see a reduction here. Other opinions welcome. [7]
Discount for broken gear:
Guys - we need a solid approach for this one. Do we have an engineer/ expert here who can help out? Some number like "1 out of 100 chips will fail within the first 60 days". Any hint is valuable as I think we need to include and maintain this position. [8]
Punching the number in our spreadsheet we have the following new picture (see below).
Now the hard part. Setting a new mid-Tier forecast value. In comparison to our first shot we have two major developments that need to be adressed.
A) Massive new hash power from various known and unknown sources (7.7 PH alone in January)
B) Basically all major new vendors are late, struggling or close to bust
Conclusion: Mainly based on the massive amount of hashing power added by various forces which is partly but not completely compensated by negative vendor behaviour - I will increase the Forecast by +27 PH to 95.000 Thash (13.3b diff) as of 01.07.2014. Basically I estimate we will see roughly 10-15 PH added to the network from now on until July.
Next update will be after 3-5 jumps with new basis or in case something major happens.
Thanks for following and providing valuable information!
Sources:
[1] No obvious major vendor shipped between 01.01.2014 and 05.02.2014. Roughly 8 PH got added simply from "Other" sources
[2] https://www.kncminer.com/news
[3] Following https://bitcointalk.org/index.php?topic=99497.17502
[4] Following https://bitcointalksearch.org/topic/hashfast-announces-specs-for-new-asic-400ghs-262052
[5] Following https://bitcointalksearch.org/topic/cointerra-announces-its-first-asic-hash-rate-greater-than-500-ghs-269093
[6] Following https://bitcointalk.org/index.php?topic=291141.1280
[7] Following https://bitcointalksearch.org/topic/black-arrow-28nm-100ghash-bitcoin-asic-from-049ghs-294197
[8] No source yet. Any ideas welcome
--------------------------------------------------------------------------------------------------------------------
Update 07.04.2014
Final Words and Closing
Everybody, this is my last update on this thread. I am quite bound in different other tasks and will not have the time to observe, analyze and comment all current trends in detail anymore. This is unfortunate but sloppy analysis due to time pressure is below my standards, so I better quit to not give any false predictions - Thank you everybody for your inputs and useful hints. Feel free to continue the thread or overtake.
This said, it doesnt mean I completely abandon our forecast of ~95 PH by end of June 2014. In contrary, I think it is still very viable even without looking forward on the last three months of the observed timeframe.
Why? Our last datapoint is from Feb 05 2014 with a hashrate of 18.8 PH and last jump is from Apr 05 2014 with 43.8 PH which means 2 month account for roughly 25 PH. In our last forecast update we predicted 10-15 PH per month until July. Looking on the last two months isolated our prediction was quite accurate until now and I dont see any game changing reasons.
Right now we have different trends that slow down and reduce the PHs added (low BC price, MTGox damage, massive delays of basically all major vendors, rising energy expenses), but also compensating effects which speed up the growth (mainly new vendors popping up every day, lower hardware prices, massive datacenters etc...). This all evens out somehow and adds up to something between 10-15 PH per month for the next 3 months.
With big caution - final prediction: 90-100 PH by end of June 2014.
Cheers & I am out
Basically, this thread is a continuation of DeathAndTexas's thread using the scheduled/expected delivery date of all main ASIC players to assemble a somewhat valid forecast of total network hashrate. Personally, I think this is the most appropiate way to do it and I highly appreciate his work. I will also copy the basic design of his thread - it is in good shape and people involved earlier will recognize it fast. I was hoping he would continue updating his thread on a regular basis - this said, I have no intention in stealing anyones idea - I am just interested in a valid hashrate forecast. Mr DeathAndTexas - in case I spit in your soup here - feel free to drop me a PM. Link to original thread -> Original thread
Goal
Produce a hashrate forecast
Method
A) In case you have an indication/ wild speculation/ backed up source of how much Thash will be released by a specific entity post your estimate in the following format:
Code:
Thash: xyz Thash
Origin: Random Mining Company
Product: Megasuperhasher Batch I
Expected Delivery: Until Q1/ Until Q2
Tier: Lower/ Mid/ Upper
Source: www.xyz.de or own estimate
Reasoning: I believe Random Mining Company will drop xyz Thash...in Q1 2014.... due to... blablabla....
B) Any estimate is welcome. In case you want to contradict a statement please behave mannerful and explain why your estimate differs
C) Always try to include a source and a statement about your scenario assumptions 1) Lower tier, 2) Mid tier, 3) Upper tier
D) This thread is not to discuss any specific products in terms of ROI. Whether Miner xyz will break even is offtopic and will be deleted
E) Any post containing unrespectful behavior or other offtopic discussions will be deleted
F) Before posting check out the scam list first! -> Good effort!
G) I am open for any feedback or ideas how to improve our forecast approach. Don't post it here. Leave me a PM.
H) No one can predict the future. Let's try to generate a realistic case. In the end we will most probably have a range of xxx-xxx Thash.
Total Hashrate Forecast Q2 2014
Based on delivery by end of Q2 2014:
Code:
Starting Point 01.01.2014 11.000 Thash
Lower Mid Upper Source/ Comment
KnC (batch 1, Sold out) 3.600 4.200 4.800 [1]
KnC (batch 2, Sold out) 3.600 4.200 4.800 [2]
KnC (batch 3, Selling right now) 3.600 4.200 4.800 [3]
KnC (Potential additional Jupiter Batch) 1.200 2.000 2.800 [4]
Asicminer (no information on business approach) 2.000 10.000 20.000 [5]
Hashfast (Batch "Oct", Sold out) 220 275 330 [6]
Hashfast (Batch "Oct" MPP) 660 825 990 [7]
Hashfast (Batch "Nov/ Dec", Sold out) 1.000 2.000 2.500 [8]
Hashfast (Batch "Nov/ Dec" MPP) 3.000 6.000 7.500 [7]
Hashfast (Batch 3, Sold out) 1.000 1.500 3.000 [8]
Hashfast (Batch 4, Selling right now) 1.000 1.500 3.000 [8]
Hashfast (Batch 5, not open yet) 1.000 1.500 3.000 [8]
Hashfast (Batch 6, not open yet) 1.000 1.500 3.000 [8]
Cointerra (Batch 1, Dec, Sould out) 1.000 1.200 1.200 [9]
Cointerra (Batch 2, Jan, Sold out) 1.000 1.200 1.200 [9]
Cointerra (Batch 3, Feb, Sold out) 1.000 1.200 1.200 [10]
Cointerra (Batch 4, March, Sold out) 1.000 1.200 1.200 [10]
Cointerra (Batch 5, April, Sold out) 1.000 1.500 3.000 [10]
Cointerra (Batch 6, May, Selling righ now) 1.000 1.500 3.000 [10]
Cointerra (Batch 7, June, not open yet) 1.000 1.500 3.000 [10]
Bitmine ("Until March batch") 3.000 3.500 4.000 [11]
Bitmine ("Follow up batches until Q2") 3.500 4.000 6.000 [12]
BFL ("Monarch") 10.000 20.000 40.000 [13]
Blackarrow (Batch 1) 800 800 1.200 [14]
Blackarrow (Batch 2) 800 1.000 1.200 [15]
Blackarrow (Batch 3) 800 1.000 1.200 [15]
Secret Mining Operation & Others 2.000 5.000 10.000 [16]
Discount for broken gear -1500 -1.000 -800 [17]
-----------------------------------------------------------------------------------------
Forecast as of 01.07.2014 (raw) 49.280 83.300 137.420
Forecast as of 01.07.2014 (adjusted) 40.000 68.000 170.000
Conclusion, Chances & Sensitivities (Preliminary)
Rather than just posting a number I prefer to lead the discussion based on a range or tier system. For the first draft of this conclusion I introduce the lower, mid and upper tier.
Lower tier:
Should be the number we confidently can assume as the minimum hash rate as of 01.07.2014. Lets have a look - our first drafted numbers show a hash rate of around 50 Ph and represent what would happen in case all major players slightly underdeliver. Honestly, this is high risk, high pressure start up area. Failure & delay will happen. Moreover, I expect at least 1 (2 is more likely) major players (BFL, KnC, Cointerra, Hashfast, Bitmine, Asicminer, Bitfury, Avalon, Blackarrow) to face massive problems in delivery or simply go bust.
Following this - for the first draft - I set Lower tier to ~40 Ph (~5.6b diff)
Following this - for the first draft - I set Lower tier to ~40 Ph (~5.6b diff)
Upper tier:
Should be the number we confidently can assume as the maximum hash rate as of 01.07.2014. Basically, the maximum hashrate includes the assumption that all major players squeeze their production capacity, produce day and night, never sleep, have a perfect supply chain, find enough people to buy and run their products and never do any management mistakes. We all know this will not happen. I also don't think we will face the "endgame" scenario within the next 6 months, but this graph shows where we will end up assuming customers are ready to buy in a 12 month break-even window.
Following this - although its hard to draft a number - I set Upper tier to ~170 Ph (~23.8b diff)
Following this - although its hard to draft a number - I set Upper tier to ~170 Ph (~23.8b diff)
Mid tier:
Should be the number we use for investment purpose. This is the hard part. Right now we still have a lot of unknown variables in our figures. We need to continue looking for valid sources of the batch sizes of all major players - Most is still speculation right now. This said, assuming that some players will do better than expected and others compensate with massive failure I estimate a hashrate of ~45-90 Ph which means = The network grows around 10 Ph per month starting in January with - given "stable" BC/USD rate - increasing speed for sure.
Following this - for the first draft - I set Mid tier to ~68 Ph (9.5b diff)
Following this - for the first draft - I set Mid tier to ~68 Ph (9.5b diff)
First draft - Feel free to discuss this openly with your insights! Still have the gut feeling our Mid tier is too high. I expected something around 7b diff earlier.
Version
- 27.12.2013; start of discussion
- 30.12.2013; first rough draft; removed Q1 FC; set focus to Q2 FC; renamed scenarios to "Lower, Mid, Upper"; draw first conclusion
- 07.01.2013; Hashfast failed to provide, but for now -until clarifiaction of situation - lets keep FC like it is; Some unknown source keeps adding massive hashpower; We have to observe that trend; Next review around end of January; By then, we will have a lot more information for Asicminer, Bitmine, Hashfast, Cointerra etc...
Next Step/ To dos until 07-09.02.2014
- Check batch definition of each vendor - I am not 100% sure whether I did mistakes (by: Community)
- Check & discuss lower/ mid/ upper figures (by: Community)
- Ongoing search for every tiny bit of information (from mails, forum, vendor hp, IRC, media) to verify our data (by: Community)
- Update tiers & conclusion including aboves findings (by: author)
Sources for forecast data
[1] https://www.kncminer.com/news/news-69
[2] https://www.kncminer.com/news/news-72
[3] https://www.kncminer.com/categories/miners
[4] https://www.kncminer.com/news/news-73 -> Batch size is pure speculation. Any ideas?
[5] Statement Friedcat in Asicminer thread on bitcoin.org -> Someone has the link?
[6] 550 Units@ 400-600 -> HF Order Chain information; Speed still not verified
[7] max. 400% of BJ speed-> Hashfast ToS
[8] No figure available; Is currently selling batch 4, which is guaranteed to ship by March 31, 2014; Once product is debugged additional batches will follow for sure
[9] http://www.coindesk.com/cointerra-cuts-price-of-terraminer-iv-bitcoin-mining-rig/ -> The 2 Ph is Dec and Jan batch???
[10] No figure availabe; following batch 1; CT is currently selling batch 5; Cointerra has without a doubt massive capacities; any ideas?
[11] http://www.coindesk.com/bitmine-to-drop-4phs-of-asic-power-onto-bitcoin-network-before-april/
[12] Pure speculation on basis of production capacity; I read they have capacity of around 50 rigs/ day, 20 working days, 2 TH biggest unit = 2 Ph/ month
[13] Pure speculation; No one knows right now. Josh claims capacity of 1000/ day, 20 working days, 600 Gh unit = 12 Ph/ month; From the past - Company has reputation of beeing late and overstating its capability. Nevertheless, they already deployed substantial hashing power & future potential should not be underestimated
[14] http://ecointalk.net/topic/11-prepaying-full-or-50/?p=17
[15] Any information??? Which batch is selling right now? How many units?
[16] Cumulation of smaller vendors; hidden Island operations; Start-ups whatever; safety margin
[17] Broken gear; numbers wild speculation
Let's do this.
--------------------------------------------------------------------------------------------------------------------
Update 11.02.2014
Finally I found some time to scramble the numbers. We got a lot of new information and developments. I try to cover them all. In case I miss something crucial feel free to give me the hint. In addition - before I start with this - at some point I will surely drop negative/ positive remarks about vendor x or y. This is not to attack/ promote the operation or the people behind it. This is simply the attempt to transfer their behaviour into some viable numbers. Fans of club A or B. Please do not feel judged - try to see this in the right light: We want a solid FC here!
Let's start. First - I want to check what happened between our starting point 01.01.2014 (11 PH) and the last adjustment on 05.02.2014 (18.7 PH). In this period roughly +8/9 PH got added to the network. Our initial guess of 10 PH per month up until July was surprisingly accurate for the first month. We all can agree that no major player in our spreadsheet shipped significant amounts. We saw some Hashfast/ Cointerra units in the wild but major consensus is that they have not started mass production yet.
Conclusion: All added network power originates from the position "Others" in our spreadsheet. Who are the players in here? We hear a few names. Most prominent should be Avalon chips (6 PHs sold), Bitfury, Antminer, 21e something... and maybe a few other hidden projects. Bad news is we have no deeper insights on the developments here. Good news is we dont need all the details to adjust our FC accordingly. It should be clear by now that my initial number of "Others" was WAY, WAY, WAAAAY to low. Given around 8 PH originates from "Others" in January I set FC to approx. 5-10 PH per month from now on only for this position. [1]
Let’s check step by step what our main spreadsheet positions are doing. From top to bottom.
KnC:
Will not deliver any additional Jupiters. Maybe a small batch of upgrade modules which will not have significant effects on the hashrate. They started to build an arctic mega-datacenter on the dark side of the Neptune to cover potential losses for their 20nm customers and move towards cloud based mining in the future. Nevertheless, they claim to have made significant progress on their development. I will only remove the Jupiter position in the spreadsheet and - for now - leave the "Neptune-Batch" position untouched. My feeling is, they either provide the physical Neptune in Q2 (unlikely) or the equivalent in virtual hashpower (more likely). Anyway, this does not impact our FC. I know there are people out there with different opinions about what KnC will do or not do. Still open for discussion here. [2]
Asicminer:
The quote of 2 to 20 PH for the initial batch of their new chip is still up to date BUT - they also announced to delay their tape-out to mid Feb. I am not an engineer - but my understanding here is: From tapeout to working chip around 60 days (mid April). Then they have to sell their chips, ship it, people need to start developing miners based on their chips and so on..... From observing their thread I also have the strong feeling their management lacks basic customer/ shareholder care, fast decision making and clear strategy. For me these are strong indicators for a delay. Although their chip looks very potent on paper - I just dont see any significant volume from AM in our observed timeframe. This said, AM FC is reduced significantly. [3]
Hashfast:
I guess not much to say here. They messed it up big time. As far as I followed the 1m+ pages in their tread they managed to ship their initial batch without MPP. Don't know whether they started batch 2. Clear is: They have engineering problems, they have court action going on from changing their ToS several times after sale and they suffered BIG reputation damage. Basically, they are out of business or very close. Only a brain dead monkey would buy from them in future. Maybe they recover, maybe not. Big potential to reduce FC. As of now - I will remove batch 5&6 from spreadsheet. Let's wait and see what they will do in Feb/ March to judge further. Maybe they mine with own chips? We simpy do not know. [4]
Cointerra:
These guys were said to have the biggest potential of the new vendors showing up. Solid staff, solid financials and overall profesional appearance. Fact is they are also late on their initial batch. As far as I followed their thread they are still struggeling to ship out their December units. We mainly see some guys shaking hands with the CEO for marketing purpose - no mass volume. Nevertheless, first units appear to be in hand by respected members of the community.
What else? Problems (challenges) with certification, problems with performance of the unit, problems with power consumption, problems with the boards. They have a working product: Yes. They do not have it shipped in masses and clearly under the expected performance. I still think these guys will pull it off and deliver what they promise soon. But I will reduce the FC due to the 20% less performance of the units. [5]
Bitmine:
Hmmm - I clearly have no idea. They are late yes. There seem to be some working chips yes. Do they seem to be profesional enough to ship in large quantities? - maximum chance 50% chance imho. Behaviour from the last 1-2 months: Strong lack of transparency and delaying tactics. They are buying time. Their customer base starts to get nervous. Let's see and wait. FC unchanged. [6]
BFL
Just dont want to comment on these guys. I will keep FC like it is. Next time maybe. Still see significant space for reduction here. The monarch looks like it will turn to molten lava in the second you power it....
blackarrow:
Another company basically predestined to fail. Horrible customer support. Last minute delay from February to MAI for their initial batch. Their chips/ product will be worthless by then. Customers start to jump the sinking ship. Chance of going bust within 6 month: Very high. Just my judgement. Anyway, the shipping volume we have in our spreadsheet is very low. I will let it unchanged this time. Next time we may see a reduction here. Other opinions welcome. [7]
Discount for broken gear:
Guys - we need a solid approach for this one. Do we have an engineer/ expert here who can help out? Some number like "1 out of 100 chips will fail within the first 60 days". Any hint is valuable as I think we need to include and maintain this position. [8]
Punching the number in our spreadsheet we have the following new picture (see below).
Code:
Starting Point 01.01.2014 11.000 Thash
+ Ph added 7.700 Thash (mainly "Others")
Checkpoint I 05.02.2014 18.700 Thash
Lower Mid Upper Source/ Comment
KnC (batch 1, Sold out) 3.600 4.200 4.800 [2]
KnC (batch 2, Sold out) 3.600 4.200 4.800 [2]
KnC (batch 3, Sold out) 3.600 4.200 4.800 [2]
Asicminer (want to sell chips mainly) 2.000 5.000 15.000 [3]
Hashfast (Batch "Oct", Sold out) 220 275 330 [4]
Hashfast (Batch "Oct" MPP) 660 825 990 [4]
Hashfast (Batch "Nov/ Dec", Sold out) 1.000 2.000 2.500 [4]
Hashfast (Batch "Nov/ Dec" MPP) 3.000 6.000 7.500 [4]
Hashfast (Batch 3, Sold out) 1.000 1.500 3.000 [4]
Hashfast (Batch 4, Selling right now) 1.000 1.500 3.000 [4]
Cointerra (Batch 1, Dec, Sould out) 800 960 960 [5]
Cointerra (Batch 2, Jan, Sold out) 800 960 960 [5]
Cointerra (Batch 3, Feb, Sold out) 800 960 960 [5]
Cointerra (Batch 4, March, Sold out) 800 960 2.400 [5]
Cointerra (Batch 5, April, Sold out) 800 960 2.400 [5]
Cointerra (Batch 6, May, Selling righ now) 800 960 2.400 [5]
Cointerra (Batch 7, June, not open yet) 800 960 2.400 [5]
Bitmine ("Until March batch") 3.000 3.500 4.000 [6]
Bitmine ("Follow up batches until Q2") 3.500 4.000 6.000 [6]
BFL ("Monarch") 10.000 20.000 40.000
Blackarrow (Batch 1) 800 800 1.200 [7]
Blackarrow (Batch 2) 800 1.000 1.200 [7]
Blackarrow (Batch 3) 800 1.000 1.200 [7]
Secret Mining Operation & Others 25.000 37.500 50.000 [1]
Discount for broken gear -1500 -1.000 -800 [8]
-----------------------------------------------------------------------------------------
Forecast as of 01.07.2014 (raw) 86.380 121.920 180.700
Forecast as of 01.07.2014 (adjusted) 50.000 95.000 170.000
Now the hard part. Setting a new mid-Tier forecast value. In comparison to our first shot we have two major developments that need to be adressed.
A) Massive new hash power from various known and unknown sources (7.7 PH alone in January)
B) Basically all major new vendors are late, struggling or close to bust
Conclusion: Mainly based on the massive amount of hashing power added by various forces which is partly but not completely compensated by negative vendor behaviour - I will increase the Forecast by +27 PH to 95.000 Thash (13.3b diff) as of 01.07.2014. Basically I estimate we will see roughly 10-15 PH added to the network from now on until July.
Next update will be after 3-5 jumps with new basis or in case something major happens.
Thanks for following and providing valuable information!
Sources:
[1] No obvious major vendor shipped between 01.01.2014 and 05.02.2014. Roughly 8 PH got added simply from "Other" sources
[2] https://www.kncminer.com/news
[3] Following https://bitcointalk.org/index.php?topic=99497.17502
[4] Following https://bitcointalksearch.org/topic/hashfast-announces-specs-for-new-asic-400ghs-262052
[5] Following https://bitcointalksearch.org/topic/cointerra-announces-its-first-asic-hash-rate-greater-than-500-ghs-269093
[6] Following https://bitcointalk.org/index.php?topic=291141.1280
[7] Following https://bitcointalksearch.org/topic/black-arrow-28nm-100ghash-bitcoin-asic-from-049ghs-294197
[8] No source yet. Any ideas welcome
--------------------------------------------------------------------------------------------------------------------
Update 07.04.2014
Final Words and Closing
Everybody, this is my last update on this thread. I am quite bound in different other tasks and will not have the time to observe, analyze and comment all current trends in detail anymore. This is unfortunate but sloppy analysis due to time pressure is below my standards, so I better quit to not give any false predictions - Thank you everybody for your inputs and useful hints. Feel free to continue the thread or overtake.
This said, it doesnt mean I completely abandon our forecast of ~95 PH by end of June 2014. In contrary, I think it is still very viable even without looking forward on the last three months of the observed timeframe.
Why? Our last datapoint is from Feb 05 2014 with a hashrate of 18.8 PH and last jump is from Apr 05 2014 with 43.8 PH which means 2 month account for roughly 25 PH. In our last forecast update we predicted 10-15 PH per month until July. Looking on the last two months isolated our prediction was quite accurate until now and I dont see any game changing reasons.
Right now we have different trends that slow down and reduce the PHs added (low BC price, MTGox damage, massive delays of basically all major vendors, rising energy expenses), but also compensating effects which speed up the growth (mainly new vendors popping up every day, lower hardware prices, massive datacenters etc...). This all evens out somehow and adds up to something between 10-15 PH per month for the next 3 months.
With big caution - final prediction: 90-100 PH by end of June 2014.
Cheers & I am out
Jump to: