490,000 Avalon chips already ordered - 150T hashrate spike coming in August - page 4.

minternj

sr. member

Activity: 434

Merit: 250

Quote from: Transisto on April 22, 2013, 11:38:22 AM

What indicate 490k chips ? The five 3000Btc orders aren't factors of 782btc.

I thought the same but if you follow the transaction infro from addresses they are orders of ~15k BT which is 2 orders of chips. Someone pointed this out and I now agree that is all chip orders in that address.

Transisto

donator

Activity: 1731

Merit: 1008

What indicate 490k chips ? The five 3000Btc orders aren't factors of 782btc.

Dalkore

legendary

Activity: 1330

Merit: 1026

Mining since 2010 & Hosting since 2012

Quote from: bcpokey on April 22, 2013, 03:09:09 AM

I think what that earlier fellow meant was that pricing and difficulty will be relatively in sync when actors behave rationally. It seems that this asic phase is akin to the 2011 gpu phase when bitcoin price exploded, in that people acted foolishly. Just as people were taking out $20k loans to retrofit their house and buy GPUs and rigs (that they later regretted), we now see people pricing themselves in at $50,000 for a batch 1 avalon, $9000 for an ASICMiner blade, $80,000 for a batch of avalon chips, rather willy nilly. If this thread is correct, and 150TH of chips has been ordered *already* and people have shown little sign of slowing down on that unlimited train, there might be another glut of "oops, I spent way way way more than I should have" rolling down the track.

Would not surprise me if we once again surpass any reasonable plateau, and find ourselves in deep waters. A shame too, because it makes it very difficult to participate when I can't count on the other actors to not slit my throat along with their own.

Well put. Yes, I believe we are in "deep waters" so you will need to be really patient to ROI. One interesting side-effect is that now the network will be more secure for a longer period of time because miners want to recoup their investments. This creates an interesting situation. With the ROI being pushed out this far, I will be looking to expand my management service into a very very cheap power district much sooner than I thought to allow people to send their machines to me so their ROI time-frame improves. Road-trip time.

Signus

newbie

Activity: 28

Merit: 0

You can look at an estimate for loss of profit based off of the expected rise in difficulty from now until about the 1st of August.

I did the analysis for a 47.3GH/s FPGA rig as an example. An ASIC rig would just have a lower initial cost but the profit degradation would be the same.

Look for my post here: https://bitcointalksearch.org/topic/m.1910302

organofcorti

donator

Activity: 2058

Merit: 1007

Poor impulse control.

Quote from: qbits on April 22, 2013, 04:36:12 AM

Quote from: organofcorti on April 21, 2013, 08:01:42 PM

Quote from: qbits on April 21, 2013, 07:58:44 PM

Quote from: organofcorti on April 21, 2013, 07:37:47 PM

Quote from: qbits on April 21, 2013, 04:40:33 PM

Unlike with gpu asic has not other useful purpose. If diff is too high, people will get rid of them and never come back. Consolidation of hash power is ineviteable.

I'm not following you. I say that mining difficulty correlates strongly with past BTCUSD prices. Are you saying this is or is not the case?

here is what i think:
gpu has other purposes besides btc mining. if marginaly profitable i'll mine. if not i'll stop. at worst sell off a card or two. but i can always be back if need be. so yes, current diff is based on past price.

asic is different. i'll buy it to mine period. if unprofitable i'll sell it. but it's likely that to be profitable i need to run it like a business. large scale because of upfront cost. if in this scenario i loose interest in mining then i'll sell hw and whole company to a competitor. and will not be back.

with asic it's a one way street. and will result in small number of dominant players.

And so you do or you don't think difficulty will correlate with the exchange rate?

as i said:
- with gpu: price drives diff (price up is followed by hash up, price down hash down)
- with asic: price up hash up, price down hash remains and miners consolidate i.e. there are fewer but larger miners

Ah, thanks. Sometimes I need my information pre-chewed Wink

I agree to a point, but I think that electricity prices will have a significant effect as well.

Edit: What you're suggesting already happened after the last exchange rate bubble, so it's not specific to ASICs, although the effect may be greater.

qbits

full member

Activity: 219

Merit: 100

Quote from: organofcorti on April 21, 2013, 08:01:42 PM

Quote from: qbits on April 21, 2013, 07:58:44 PM

Quote from: organofcorti on April 21, 2013, 07:37:47 PM

Quote from: qbits on April 21, 2013, 04:40:33 PM

Unlike with gpu asic has not other useful purpose. If diff is too high, people will get rid of them and never come back. Consolidation of hash power is ineviteable.

I'm not following you. I say that mining difficulty correlates strongly with past BTCUSD prices. Are you saying this is or is not the case?

here is what i think:
gpu has other purposes besides btc mining. if marginaly profitable i'll mine. if not i'll stop. at worst sell off a card or two. but i can always be back if need be. so yes, current diff is based on past price.

asic is different. i'll buy it to mine period. if unprofitable i'll sell it. but it's likely that to be profitable i need to run it like a business. large scale because of upfront cost. if in this scenario i loose interest in mining then i'll sell hw and whole company to a competitor. and will not be back.

with asic it's a one way street. and will result in small number of dominant players.

And so you do or you don't think difficulty will correlate with the exchange rate?

as i said:
- with gpu: price drives diff (price up is followed by hash up, price down hash down)
- with asic: price up hash up, price down hash remains and miners consolidate i.e. there are fewer but larger miners

bcpokey

hero member

Activity: 602

Merit: 500

I think what that earlier fellow meant was that pricing and difficulty will be relatively in sync when actors behave rationally. It seems that this asic phase is akin to the 2011 gpu phase when bitcoin price exploded, in that people acted foolishly. Just as people were taking out $20k loans to retrofit their house and buy GPUs and rigs (that they later regretted), we now see people pricing themselves in at $50,000 for a batch 1 avalon, $9000 for an ASICMiner blade, $80,000 for a batch of avalon chips, rather willy nilly. If this thread is correct, and 150TH of chips has been ordered *already* and people have shown little sign of slowing down on that unlimited train, there might be another glut of "oops, I spent way way way more than I should have" rolling down the track.

Would not surprise me if we once again surpass any reasonable plateau, and find ourselves in deep waters. A shame too, because it makes it very difficult to participate when I can't count on the other actors to not slit my throat along with their own.

peetah

member

Activity: 364

Merit: 10

Quote from: Dalkore on April 22, 2013, 01:53:21 AM

Quote from: peetah on April 22, 2013, 12:15:23 AM

I think Avalon started something very interesting in terms of your chart. And that is they removed USD from the equation by pricing in BTC. As such, until they stop doing that, or competition forces them to price in fiat, there may be a duration where the exchange rate does not dictate the proliferation of miners. This is double the case since asicminer by default is denominated in BTC. Arguably tripled as well as it seems BFL will price in reaction to Avalon. The only price unlinkable to fiat is electricity, but currently that is not the main concern of asic miners.

So what your saying is because Avalon & ASICMiner are priced only in BTC, they are retaining a certain portion of BTC from miners forced upgrade to ASIC (assume they are keeping a BTC reserve in a form of profits)? That would make sense because the nature of ASIC mining, it is a continual race toward efficiency until (as organofcorti said) we hit the hash-rate plateau.

Actually no that wasn't my point. But your point is much more valid. Since asics can only produce BTC, even asic producers are forced to think in BTC and not fiat. Both asicminer and Avalon hold more BTC than fiat. So their decision to take short term profit or invest in new technology is also based on BTC. Furthermore, upgrades to asic technology will only produce BTC. As their payback period is even longer than miners, it is arguable that this BTC based decision making is here to stay.

Dalkore

legendary

Activity: 1330

Merit: 1026

Mining since 2010 & Hosting since 2012

Quote from: peetah on April 22, 2013, 12:15:23 AM

I think Avalon started something very interesting in terms of your chart. And that is they removed USD from the equation by pricing in BTC. As such, until they stop doing that, or competition forces them to price in fiat, there may be a duration where the exchange rate does not dictate the proliferation of miners. This is double the case since asicminer by default is denominated in BTC. Arguably tripled as well as it seems BFL will price in reaction to Avalon. The only price unlinkable to fiat is electricity, but currently that is not the main concern of asic miners.

So what your saying is because Avalon & ASICMiner are priced only in BTC, they are retaining a certain portion of BTC from miners forced upgrade to ASIC (assume they are keeping a BTC reserve in a form of profits)? That would make sense because the nature of ASIC mining, it is a continual race toward efficiency until (as organofcorti said) we hit the hash-rate plateau.

organofcorti

donator

Activity: 2058

Merit: 1007

Poor impulse control.

Quote from: peetah on April 22, 2013, 12:15:23 AM

I think Avalon started something very interesting in terms of your chart. And that is they removed USD from the equation by pricing in BTC. As such, until they stop doing that, or competition forces them to price in fiat, there may be a duration where the exchange rate does not dictate the proliferation of miners. This is double the case since asicminer by default is denominated in BTC. Arguably tripled as well as it seems BFL will price in reaction to Avalon. The only price unlinkable to fiat is electricity, but currently that is not the main concern of asic miners.

I think you're partly right - for the moment the exchange rate will have little effect on the profitability (edit: vs unprofitability) of ASIC miners. However, the nature of the difficulty changes means that we will quickly advance to the point that electricity costs will be a problem for ASIC miners too

At this point the relationship between exchange rate and difficulty wil be much more obvious and at that point we hit a hashrate plateau until the next big change in either exchange rate or mining device design.

peetah

member

Activity: 364

Merit: 10

I think Avalon started something very interesting in terms of your chart. And that is they removed USD from the equation by pricing in BTC. As such, until they stop doing that, or competition forces them to price in fiat, there may be a duration where the exchange rate does not dictate the proliferation of miners. This is double the case since asicminer by default is denominated in BTC. Arguably tripled as well as it seems BFL will price in reaction to Avalon. The only price unlinkable to fiat is electricity, but currently that is not the main concern of asic miners.

organofcorti

donator

Activity: 2058

Merit: 1007

Poor impulse control.

Quote from: jspielberg on April 21, 2013, 10:25:20 PM

Hey Organofcorti,

The "Price drives Difficulty" thread, and your blog spot article, are interesting. Numbers don't lie and your model with R^2=0.9991 is hard to dispute.

Philosophically though I think it is a shame. It sort of implies that casual mining is a fools errand, and resources are better spent just acquiring BTC rather than mining. Personally I find that boring... but I guess now I have been warned.

In the meantime I will stop saying the sun revolves around the earth Grin

I think you're looking at it the wrong way and that's probably my fault. The exchange rate drives the network hashrate through the cost of electricity, returns on equipment and so on. It illustrates how the exchange rate affects miners directly.

There are other drivers that can't be accounted for by the model and we'll be seeing plenty of these soon.

Up until now though the main drivers of the network hashrate has been previous exchange rates and previous network hashrates.

organofcorti

donator

Activity: 2058

Merit: 1007

Poor impulse control.

Quote from: peetah on April 21, 2013, 10:25:16 PM

Organofcorti: that is an amazing chart. How can correlation be so close?

Thanks for the complement. If you want more information, click on the chart and the link will take you to the post that describes how it's done. You might want to start at the previous "10.1" post for some background.

Quote from: peetah on April 21, 2013, 10:25:16 PM

But surely that is only the case for red sea competition. More fish in the sea, more fishing boats. But asic is blue ocean. Is there a modelling method that you can apply to predict the next 18 months? Say when gpu's or fpga's first hit the scene?

The modelling method survived the start of GPUs and FPGAs quite well. I don't know how it's going to go over the next few years, but I'm keeping track. If you search my blog for "Weekly network forecast" you can see the most recent updates. At the moment with a volatile USD BTC rate and ASICs arriving in spurts the predictive value isn't as good as it was a few weeks ago.

jspielberg

sr. member

Activity: 490

Merit: 255

Hey Organofcorti,

The "Price drives Difficulty" thread, and your blog spot article, are interesting. Numbers don't lie and your model with R^2=0.9991 is hard to dispute.

Philosophically though I think it is a shame. It sort of implies that casual mining is a fools errand, and resources are better spent just acquiring BTC rather than mining. Personally I find that boring... but I guess now I have been warned.

In the meantime I will stop saying the sun revolves around the earth Grin

peetah

member

Activity: 364

Merit: 10

Organofcorti: that is an amazing chart. How can correlation be so close? But surely that is only the case for red sea competition. More fish in the sea, more fishing boats. But asic is blue ocean. Is there a modelling method that you can apply to predict the next 18 months? Say when gpu's or fpga's first hit the scene?

organofcorti

donator

Activity: 2058

Merit: 1007

Poor impulse control.

Quote from: Dalkore on April 21, 2013, 09:02:28 PM

Quote from: jspielberg on April 21, 2013, 08:10:59 PM

I don't think difficulty is related to the exchange rate.
The algorithm dictates a steady flow of 50400 BTC every two weeks regardless of the amount of processing power. Processing power just decides how those 50,000 coins are distributed.

Public exposure I believe to the be dominant component with exchange rate.

There is a good case for this. Short-term, there might be light correlation but overall, growth of the network and demand will be the ultimate driving factor.

The original quote was :

Quote from: Viceroy on April 21, 2013, 04:27:59 PM

Quote from: qbits on April 21, 2013, 04:27:03 PM

What do you think will happen with btc price in august?

This question is unanswerable and belongs in economic speculation. There is no correlation between mining difficulty and the price of the coin, price is determined by human demand which is in-calculate-able.

This is an answer about the past, not the future. Up until now there has been an extreme clear correlation between Difficulty and previous prices.

One such linear model (for hashrate rather than difficulty, but the same principle applies):

Code:

log(H) ~ 1.74 + 0.94lag1(log(H)) + 0.21lag1(log(p)) - 0.14lag4(log(p)), CI = +/- 15.1%

where H = network hashrate, p = BTCUSD price.

Liner model compared to network hashrate:

Dalkore

legendary

Activity: 1330

Merit: 1026

Mining since 2010 & Hosting since 2012

Quote from: jspielberg on April 21, 2013, 08:10:59 PM

I don't think difficulty is related to the exchange rate.
The algorithm dictates a steady flow of 50400 BTC every two weeks regardless of the amount of processing power. Processing power just decides how those 50,000 coins are distributed.

Public exposure I believe to the be dominant component with exchange rate.

There is a good case for this. Short-term, there might be light correlation but overall, growth of the network and demand will be the ultimate driving factor.

organofcorti

donator

Activity: 2058

Merit: 1007

Poor impulse control.

Quote from: jspielberg on April 21, 2013, 08:10:59 PM

I don't think difficulty is related to the exchange rate.
The algorithm dictates a steady flow of 50400 BTC every two weeks regardless of the amount of processing power. Processing power just decides how those 50,000 coins are distributed.

Public exposure I believe to the be dominant component with exchange rate.

"Eppur si muove"

peetah

member

Activity: 364

Merit: 10

Quote from: qbits on April 21, 2013, 07:33:10 PM

Quote from: Viceroy on April 21, 2013, 04:43:28 PM

Quote from: qbits on April 21, 2013, 04:40:33 PM

Unlike with gpu asic has not other useful purpose. If diff is too high, people will get rid of them and never come back. Consolidation of hash power is ineviteable.

Unless you use a different algorithm... See here. I agree this hashing power may spell the end of bitcoin as we know it. Disagree? Come talk to me about it.

sure. i read the suggested thread and you are correct in your analysis. from the merchant standpoint i don't care what crypto currency i use. if ltc has faster confirmation then great. nice but not spectacular.

but... if btc with diff of 500M is only worthwhile to mine by 100 entities, i might consider it to be unsafe. as if only one entity ended up mining it would certainly be unsafe.

simply put: safety of btc is in number of people mining it and using it. asic goes against that and has therefore likely killed off btc.

ltc? well if miners move there, so will exchanges. then merchants and users will surely follow.

personaly i predict btc ltc price parity within a year (1 btc = 1 ltc)

cheers...

Security is defined by any one person with the ability to takeover the network. When there are three suppliers (two and a half? ) pumping out machines that can compete against all the super computers worldwide then we will have achieved network stability on two counts. Firstly government will never catch up with the specialized hardware to compete. Secondly, machine distribution is scattered worldwide. Although this has been discounted by BFL's non delivery, but chip sales has just more than compensated. There are at least 10 different threads in five different languages calling for pooled effort of turning chips into hashing machines. As such, the supply base has already become 10 fold, to the extent that ASICminers 250T is no longer a big threat.

Conversely, LTC miners can mine LTC today, XYZTC tomorrow. Which means there's no substance in it's hashing base. Furthermore, the difficulty is still within grasps of super computers. So, LTC has yet to see the light.

jspielberg

sr. member

Activity: 490

Merit: 255

I don't think difficulty is related to the exchange rate.
The algorithm dictates a steady flow of 50400 BTC every two weeks regardless of the amount of processing power. Processing power just decides how those 50,000 coins are distributed.

Public exposure I believe to the be dominant component with exchange rate.

Topic: 490,000 Avalon chips already ordered - 150T hashrate spike coming in August - page 4. (Read 24323 times)