ASICMINER: Entering the Future of ASIC Mining by Inventing It

[canth]

no doubt, BFL is no go, Avalon is shady, at this point Bitfury and Kncminer are the only threat to AM wealth.

Dangerous assumption, for all we know Intel could launch a 10x more efficient chip tomorrow and Nvidia announce a competing product Monday morning. Screw marketing, numbers talk here and AM could get drowned out overnight if one of the big firms wants a piece of the pie but AM have experience, a head start and a rock solid reputation so my moneys on them for the long run.

EDIT: Thanks for the update

Not possible until Bitcoin market share is least 1 Trillion. Intel alone has 100X market cap of bitcoin, the potential market is simply too tiny for these companies to produce anything for.

Yep - it's just not that interesting of a market for Intel. Besides, right now their focus is on mobile phones, tablets and the like. If laptops and PCs end up being trounced by tablets and phones, then Qualcomm, nVidia and others will end up being much more problematic for Intel. "Missing out" on the relatively small crypto currency mining market will be the least of their problems. Face it, this is still a niche market with perhaps only hundreds of thousands of customers compared to cell phones, tablets with over 1bn customers.

[nubbins]

15 btc per block seems like a lot in fees. What sort of fee per transaction would that equate to?

At today's transaction amounts - aka 50K transactions per day or about 350 transactions per block this would be about .04 BTC per transaction. However, we have to presume that the number of BTC transactions will significantly increase as BTC gains more popularity. Anyone want to take a guess about how many transactions per block in 2-3 years?

21 million transactions a day, with a BTC0.0001 fee, would give you around BTC15 in fees per block.

BTC0.0005 fee, you'd need 4.2 million tx's per day

etc

[nubbins]

for reference, visa does ~150m tx's per day

[Mastergerund]

There was a block a few days ago with 11 BTC in transaction fees, but it was a fluke. Expensive typo 

[neilol]

for reference, visa does ~150m tx's per day

This is a good comparison. Imagine if we could capture even a 10th of their volume.

It wouldn't be hard either with a stabilized Bitcoin. Businesses that accept CC pmts would instantly gain 3-5% on their top line by switching, not a very hard pitch to agree to. I know there are assumptions baked in here..but there is real promise.

[metavox]

This may be bitcoin 101 that I just missed but it would be great if someone could flush out some justification to what Friedcst said about transaction fees becoming a bigger part of the reward for mining a block. Why is it that this will happen? 15 btc per block seems like a lot in fees. What sort of fee per transaction would that equate to?

We're doing a lot of extrapolation here since the number of BTC issued per block isn't set to halve again until ~ 2017. At today's transaction amounts - aka 50K transactions per day or about 350 transactions per block this would be about .04 BTC per transaction. However, we have to presume that the number of BTC transactions will significantly increase as BTC gains more popularity. Anyone want to take a guess about how many transactions per block in 2-3 years?

Do we have any idea what portion if transaction fees AM is currently collecting on each block in addition to the created btc?

A related question is how is btc designed to continue after all block have been solved? I know this is decades away but I'm curious if this isn't a problem to worry about or if this is something that will be solved eventually.

I guess I'm thinking that since the amount of rewarded btc for every block halves eventually the reward for solving a block wil be .00001 btc and miner will continue because they will be paid a transaction fee. If that's the case then eventually the transaction fees might be a decent percentage of a regular transaction. Seems a little worrisome in the long term to a currency that has a major advantage because of low transaction fees.

Yes, you can check the mined ASICMiner blocks and their fees here - http://blockchain.info/blocks/ASICMiner

You are correct.  BTC is designed to be supported solely by fees once all coins have been mined.  The idea is that as the block reward is halved, the number of transactions will have likely increased and will offset (to a degree) the reduced mining reward.  The key here is to estimate the transaction total per block and avg fee size per transaction.  For 15 BTC in fees per block, at today's numbers (randomly selected: 394 trx @ 0.2812 btc fees), we're looking at ~.0007 btc fee per trx, or 21000 transactions per block.  That's an increased transaction volume of roughly 47x to yield 15 BTC in fees.

Now, if the relative value of BTC/fiat goes up, then the fee will likely have an inversely correlated drop - BTC doubles, fee halves.  If we assume a doubling of BTC/fiat value by the next block halve, and a halving of the average fee to .00035 btc, we would need 42000 transactions per block to yield 15 BTC in fee rewards. We're getting close to a required transaction volume increase of 100x with this scenario.   I think major retailers would need to start accepting BTC to yield these numbers in either example.  The acceptance bar here is to make sure BTC is simple enough to acquire, secure, and spend that my parents could start buying things with it at Costco.

Please feel free to critique my numbers and approach, but this is my back-of-the-napkin approximation.

[nubbins]

It wouldn't be hard either with a stabilized Bitcoin.

emphasis mine.

i think tx fees in excess of BTC12.5 per block will happen before the next reward halving.

[SebastianJu]

Hm... i read people still bought BFL-Miners "last week"... i wonder how many could be "saved" when Asicminer would be so known for their Miner-Offers like BFL is...

Even though i see that friedcat would have work with it, for example explaining what website they want, speak about design and functionality im interested to see the effect.

What effect would have advertising at bitcointalk.org? What effect other internet marketing techniques?

I think it could lead to much more sells and maybe even to less poor guys thinking BFL is the one and only legit Miner-Company.

At the end only friedcat can decide such things. And it looks like its not top priority. I tend to believe friedcat thinks there are other things he can do to maximize dividends. At least he shows it all the time. But a rethink cant be bad.

[bobboooiie]

There was a block a few days ago with 11 BTC in transaction fees, but it was a fluke. Expensive typo 

I quess he did not have to wait for a long for confimations

[freedomno1]

The delay of confirmation of this week's dividends is caused by the payments of real
dividends are chained after the satoshi transactions, which are now refused by most
of the clients. We will either stop the satoshi payments, or unchain the payment in our
modified scripts.

Thanks for the update friedcat

Think I understand what this means but can someone re-confirm it
Mass sending out to multiple paying addresses is refused by most clients on the blockchain
And is that related to the transaction fee changes in the last update of the client

[canth]

There was a block a few days ago with 11 BTC in transaction fees, but it was a fluke. Expensive typo 

I quess he did not have to wait for a long for confimations

Heh. Like lightning! I saw that block and thought it odd - it makes sense that the only way that it would happen would be a typo. A gift to the world, eh?

[twmz]

Think I understand what this means but can someone re-confirm it
Mass sending out to multiple paying addresses is refused by most clients on the blockchain
And is that related to the transaction fee changes in the last update of the client

No, it's not related to having multiple payment addresses or to transaction fees.  The satoshis transaction has a lot of dust payments in it, many that are just a single satoshi (0.00000001).  The current version of bitcoin-qt/bitcoind considers such a transaction to be spam and refuses to relay it or include it in blocks.  It will only get included in blocks by miners that have customized algorithms for including transactions (and that have an algorithm that includes spam-like transactions).

Edit: and the reason this affects the dividends payment is that the transaction for the satoshis sends its left-over "change" to an address/output that is then used as the input/source-of-funds for the dividends payment.  The dividends transaction, then, will not get confirmed until after the satoshis transaction gets confirmed.

[freedomno1]

Think I understand what this means but can someone re-confirm it
Mass sending out to multiple paying addresses is refused by most clients on the blockchain
And is that related to the transaction fee changes in the last update of the client

No, it's not related to having multiple payment addresses or to transaction fees.  The satoshis transaction has a lot of dust payments in it, many that are just a single satoshi (0.00000001).  The current version of bitcoin-qt/bitcoind considers such a transaction to be spam and refuses to relay it or include it in blocks.  It will only get included in blocks by miners that have customized algorithms for including transactions (and that have an algorithm that includes spam-like transactions).

Edit: and the reason this affects the dividends payment is that the transaction for the satoshis sends its left-over "change" to an address/output that is then used as the input/source-of-funds for the dividends payment.  The dividends transaction, then, will not get confirmed until after the satoshis transaction gets confirmed.

That makes more sense thanks for clearing that up
But if its part of a bigger unit why does that cause a delay
Sending 10.00000001 bitcoin's would not be dust

[twmz]

Think I understand what this means but can someone re-confirm it
Mass sending out to multiple paying addresses is refused by most clients on the blockchain
And is that related to the transaction fee changes in the last update of the client

No, it's not related to having multiple payment addresses or to transaction fees.  The satoshis transaction has a lot of dust payments in it, many that are just a single satoshi (0.00000001).  The current version of bitcoin-qt/bitcoind considers such a transaction to be spam and refuses to relay it or include it in blocks.  It will only get included in blocks by miners that have customized algorithms for including transactions (and that have an algorithm that includes spam-like transactions).

Edit: and the reason this affects the dividends payment is that the transaction for the satoshis sends its left-over "change" to an address/output that is then used as the input/source-of-funds for the dividends payment.  The dividends transaction, then, will not get confirmed until after the satoshis transaction gets confirmed.

That makes more sense thanks for clearing that up
But if its part of a bigger unit why does that cause a delay
Sending 10.00000001 bitcoin's would not be dust

If a transaction contains any output that is "spam-like" then the latest version of bitcoin-qt will treat the entire transaction as non-standard (and refuse to relay it or include it in blocks).  It does not depend on the sum-total of the transaction, it depends on each individual payment in the transaction.

[freedomno1]

Think I understand what this means but can someone re-confirm it
Mass sending out to multiple paying addresses is refused by most clients on the blockchain
And is that related to the transaction fee changes in the last update of the client

No, it's not related to having multiple payment addresses or to transaction fees.  The satoshis transaction has a lot of dust payments in it, many that are just a single satoshi (0.00000001).  The current version of bitcoin-qt/bitcoind considers such a transaction to be spam and refuses to relay it or include it in blocks.  It will only get included in blocks by miners that have customized algorithms for including transactions (and that have an algorithm that includes spam-like transactions).

Edit: and the reason this affects the dividends payment is that the transaction for the satoshis sends its left-over "change" to an address/output that is then used as the input/source-of-funds for the dividends payment.  The dividends transaction, then, will not get confirmed until after the satoshis transaction gets confirmed.

That makes more sense thanks for clearing that up
But if its part of a bigger unit why does that cause a delay
Sending 10.00000001 bitcoin's would not be dust

If a transaction contains any output that is "spam-like" then the latest version of bitcoin-qt will treat the entire transaction as non-standard (and refuse to relay it or include it in blocks).  It does not depend on the sum-total of the transaction, it depends on each individual payment in the transaction.

Seems a little strange to be cutting out the satoshi from the total, it means miners who handle non-standard transactions will sometimes find more blocks than those who run the normal client just because of a one satoshi output in the whole transaction chain.

Wonder if that's more of a PR issue I do remember people putting some porn links in the blockchain, and changes to prevent that
I will not battle for the satoshi just wanted to know where it went
Thanks for answering twmz
And if it somehow effects the total number of bitcoin's fractional reserve capability the whole
Bitcoin has 2.1 quadrillion raw units

To surmise
If you don't include the satoshi as a standard relay it causes delays on the sending of units because the transaction for the satoshis sends its left-over "change" to an address/output that is then used as the input for the dividend payment.  

The dividends transaction, then, will not get confirmed until after the satoshis transaction gets confirmed.

Got it now just edited it to get it better
Sent you my first tip to someone for that explanation thanks again

[Jere.Jones]

I would like to see the satoshi transactions continue.  The reason why is that it would be hard, if not impossible, to determine how many shares an address owns without it.  

There have been mentions of adding up all the dividends, dividing by 400,000 to determine the per-share dividend, and then dividing the amount received by that dividend.  There are two problems with this:

• I can not determine the number of shares that an address owns without knowing all of the transactions that were used in this weeks dividend.  As it is now, I only need look at a single transaction to determine the number of shares owned by that address.
• Determining the per-share dividend becomes problematic for a non-direct-shareholder.  While it would be easy for a direct shareholder to determine the per-share dividend (because he/she would know how much he/she received and how many shares that dividend was applied to), someone else could not.  Even if you could manage to find all the dividend transactions and sum up the outputs, you still couldn't determine the per-share dividend because you don't know which of the outputs is the change (it could be obvious, it could not) and you don't know how many shares that sum is for as evidenced by the recent shareholder spreadsheet.

There have been discussions about these satoshi transactions being considered dust by the network.  Blocks with these transactions will still get relayed.  The "economic dust" change only affects unconfirmed transactions with dust.  While it is true they won't get relayed by stock bitcoind clients, it doesn't matter because ASICMINER can include them itself.  

In the end, these satoshi transactions provide value by making the system easily auditable and understandable.  I would hate to see them removed.

[vesper39]

One simple workaround would be to globally announce a designated address that will permanently hold 1 share of AM.

This gives the amount of dividends per share each week, as well as allow anyone to simply divide to figure out the number of shares another address holds.

Im indifferent in whether satoshis are sent or not. I'm sure AM team will figure out something reasonable.

[cointoss]

Where can I buy blades or shares directly from ASICMINER?

[freedomno1]

Where can I buy blades or shares directly from ASICMINER?

Usually on the forums at different times
Or Search ASIC auctions
If your a mega investor break this darn wall (Patiently waited for the market to forget then brings it back  )
Or if the market goes to 2.7 and forgets this wall that would be the time to ask for a loan lols
https://bitcointalk.org/index.php?topic=213729.40

To TheFuneral Not to certain myself but here is an example
https://bitcointalk.org/index.php?topic=193683.0

[TheFuneral]

is there a FAQ or HOW-TO for when it comes to buying auction ASIC shares from people on this board? I've been reading around but can't find much detail about the entire process. Do you guys have any links you could point me to on the process and rules?