21e6, LLC - Secretive ASIC manufacturer that raised $5 million

[RoadStress]

The more interesting thing was the 0.22 w/GH 22nm chip, 2.5 times more efficient than the latest generation miners.
If they already have that and are running them they're making good profit.

Another statement pulled out of the ass.

you might want to reconsider this comment
http://www.coindesk.com/21-intel-bitcoin-mining-strategy/

For example, 21 indicated it had been processing bitcoin transactions with what it called the "only chip" built at computing giant Intel’s foundry, touting close relationship with US computing giant Intel.

Intel factories, the documents suggested, were responsible for at least two generations of 21 bitcoin mining chips, a 0.57 w/GH 22nm FinFET chip (codenamed CyrusOne) and a 0.22 w/GH 22nm chip (codenamed Brownfield).

I was referring about the 2.5 times more efficient. Sorry for the confusion.

[1714835342]

1714835342

[lex_minutor]

The more interesting thing was the 0.22 w/GH 22nm chip, 2.5 times more efficient than the latest generation miners.
If they already have that and are running them they're making good profit.

Another statement pulled out of the ass.

you might want to reconsider this comment
http://www.coindesk.com/21-intel-bitcoin-mining-strategy/

For example, 21 indicated it had been processing bitcoin transactions with what it called the "only chip" built at computing giant Intel’s foundry, touting close relationship with US computing giant Intel.

Intel factories, the documents suggested, were responsible for at least two generations of 21 bitcoin mining chips, a 0.57 w/GH 22nm FinFET chip (codenamed CyrusOne) and a 0.22 w/GH 22nm chip (codenamed Brownfield).

I was referring about the 2.5 times more efficient. Sorry for the confusion.

Ok, not 2.5, but 2.3
Currently the most efficient miners are at around 0.5W/GH.

[RoadStress]

Ok, not 2.5, but 2.3
Currently the most efficient miners are at around 0.5W/GH.

You do realize that might be the power consumption at chip level right? It might be at wall too, but we don't know.

[alh]

Ok, not 2.5, but 2.3
Currently the most efficient miners are at around 0.5W/GH.

You do realize that might be the power consumption at chip level right? It might be at wall too, but we don't know.

Just to be even more pedantic, we don't really know the accuracy of those values, nor any good way to verify them. I am personally doubtful about their Intel relationship, but there is also no way to confirm nor refute it. In my experience, Intel is VERY secretive about what they are doing in terms of chip fabrication and their plans. In my dealings with them, when you ask the time, they hesitate to decide if it's OK to disclose that information.   

[jimmothy]

Ok, not 2.5, but 2.3
Currently the most efficient miners are at around 0.5W/GH.

You do realize that might be the power consumption at chip level right? It might be at wall too, but we don't know.

Just to be even more pedantic, we don't really know the accuracy of those values, nor any good way to verify them. I am personally doubtful about their Intel relationship, but there is also no way to confirm nor refute it. In my experience, Intel is VERY secretive about what they are doing in terms of chip fabrication and their plans. In my dealings with them, when you ask the time, they hesitate to decide if it's OK to disclose that information.   

They are located in SF (where all the tech VC guys are) so I'd say it's definitely possible they are using Intel fabs.

Powell pointed out that the "chip names" are most likely datacenter names in which case 0.22 w/gh at the transformer would be pretty damn impressive.

Though if that's the case, I wonder why they only have ~5 PH/s if they've been one step ahead of the competition in terms of efficiency at least.

[alh]

There is another thread with a raging debate regarding 21 (used to be 21e6)

https://bitcointalk.org/index.php?topic=1065077.0

The one take away I have from Bitcoin mining is that it's important to realize that massive improvements in mining capacity, by themselves aren't really helpful. If you have a 1Megawatt facility, that's producing X Petahash, when you have the option of upgrading to new technology (i.e. more efficient), it's better to lower your electricity costs and still just produce roughly X Petahash. If you just replace all your mining gear and still burn 1Mwatt but produce 2X Petahash, it doesn't last very long. In a month, the difficulty adjustment has mitigated most of your improvement. However, if you still produce X Pethash, but you only burn 500KW, you are money ahead for longer.

Maybe 21 realizes this, and has decided they can't just add massive Petahash without hurting themselves in the long run. Maybe  your stated ~5.5PH is what they though they could add.

I won't debate the actual value of San Francisco to getting Intel to fab chips for somebody. The thread mentioned above suggests (without proof), that the .22W/GH is a chip level value, not "at the transformer".

[jimmothy]

The one take away I have from Bitcoin mining is that it's important to realize that massive improvements in mining capacity, by themselves aren't really helpful. If you have a 1Megawatt facility, that's producing X Petahash, when you have the option of upgrading to new technology (i.e. more efficient), it's better to lower your electricity costs and still just produce roughly X Petahash. If you just replace all your mining gear and still burn 1Mwatt but produce 2X Petahash, it doesn't last very long. In a month, the difficulty adjustment has mitigated most of your improvement. However, if you still produce X Pethash, but you only burn 500KW, you are money ahead for longer.

That's not really true until you have a massive amount of hashrate.

5 PH/s at 0.5 w/gh earns $12,000 per day and costs $3,000 per day for electricity*. (profit = $1800/PHs/day)

5 PH/s at 0.25 w/gh would earn ~$12,000 per day and cost $1,500 per day for electricity.  ($2100/PHs/day)

10 PH/s (adding 5 PH/s to the network) at 0.25 w/gh would earn ~$23,800 per day and cost $3,000 per day for electricity.  ($2080/PHs/day)

50 PH/s at 0.25 w/gh would earn ~$106,900 per day and cost $15,000 per day for electricity. ($1838/PHs/day)

*assuming $0.05/kwh

So even if they added 45 PH/s after upgrading, they'd only be reducing their profit per PH/s by ~13%.

[Biodom]

The one take away I have from Bitcoin mining is that it's important to realize that massive improvements in mining capacity, by themselves aren't really helpful. If you have a 1Megawatt facility, that's producing X Petahash, when you have the option of upgrading to new technology (i.e. more efficient), it's better to lower your electricity costs and still just produce roughly X Petahash. If you just replace all your mining gear and still burn 1Mwatt but produce 2X Petahash, it doesn't last very long. In a month, the difficulty adjustment has mitigated most of your improvement. However, if you still produce X Pethash, but you only burn 500KW, you are money ahead for longer.

That's not really true until you have a massive amount of hashrate.

5 PH/s at 0.5 w/gh earns $12,000 per day and costs $3,000 per day for electricity*. (profit = $1800/PHs/day)

5 PH/s at 0.25 w/gh would earn ~$12,000 per day and cost $1,500 per day for electricity.  ($2100/PHs/day)

10 PH/s (adding 5 PH/s to the network) at 0.25 w/gh would earn ~$23,800 per day and cost $3,000 per day for electricity.  ($2080/PHs/day)

50 PH/s at 0.25 w/gh would earn ~$106,900 per day and cost $15,000 per day for electricity. ($1838/PHs/day)

*assuming $0.05/kwh

So even if they added 45 PH/s after upgrading, they'd only be reducing their profit per PH/s by ~13%.

The only thing that is truly relevant is to consider whether bitcoin is finished if it is mined by just four companies and their associates.
i happen to see no future for bitcoin if and when it is ONLY produced (and dumped en masse) by Bitfury, KnC, Spond/bitcoinshop and Bitmain+associates.
This makes no market and 2014-2015 bear market and super slow (if any) rise in transactions volume proves it.
How can we, current miners, continue mining when nobody will sell most current miners to us? it's just self explanatory.

maybe 21 inc is crazy and it won't work, but without this proposal, bitcoin will be dead or severely marginalized within a few years.

[alh]

The only thing that is truly relevant is to consider whether bitcoin is finished if it is mined by just four companies and their associates.
i happened to see no future for bitcoin if and when it is ONLY produced (and dumped en masse) by Bitfury, KnC, Spond/bitcoinshop and Bitmain+associates.
This makes no market and 2014-2015 bear market and super slow (if any) rise in transactions volume proves it.
How can we, current miners, continue mine when nobody will sell most current miners to us? it's just self explanatory.

maybe 21 inc is crazy and it won't work, but without this proposal, bitcoin will be dead or severely marginalized within a few years.

So for you the threshold of "decentralized" is 5 instead of 4? So far I have seen nothing to suggest that the 21Inc miners won't all be working with a single 21Inc pool, have you? I have zero reason to believe that any such gadget will be configurable to me to any degree than beyond that required to make it work within the "21 Bitcoin universe". I supply Internet and electricity, on my dime, and maybe I get something back from 21? This just distributes geographically the miners. It doesn't significantly change the dynamics any, in my opinion.

[jimmothy]

So far I have seen nothing to suggest that the 21Inc miners won't all be working with a single 21Inc pool, have you?

IIRC they specifically mentioned it would be mining on their pool in one of the articles.

[Biodom]

The only thing that is truly relevant is to consider whether bitcoin is finished if it is mined by just four companies and their associates.
i happened to see no future for bitcoin if and when it is ONLY produced (and dumped en masse) by Bitfury, KnC, Spond/bitcoinshop and Bitmain+associates.
This makes no market and 2014-2015 bear market and super slow (if any) rise in transactions volume proves it.
How can we, current miners, continue mine when nobody will sell most current miners to us? it's just self explanatory.

maybe 21 inc is crazy and it won't work, but without this proposal, bitcoin will be dead or severely marginalized within a few years.

So for you the threshold of "decentralized" is 5 instead of 4? So far I have seen nothing to suggest that the 21Inc miners won't all be working with a single 21Inc pool, have you? I have zero reason to believe that any such gadget will be configurable to me to any degree than beyond that required to make it work within the "21 Bitcoin universe". I supply Internet and electricity, on my dime, and maybe I get something back from 21? This just distributes geographically the miners. It doesn't significantly change the dynamics any, in my opinion.

It seems to me that:

1. if you have/get stream of satoshis, you will learn how to use them, then maybe want to replenish them from outside, increasing buying of bitcoin and hence at least achieving balance if not sponsoring bitcoin appreciation.
2. it is very unlikely that you would immediately dump your satoshis in comparison with the aforementioned four mining companies, which do the dumping in a heartbeat (allegedly).
3. since you already have an account, all kind of payments could be directed to you in satoshis instead of other currencies due to cheaper/easier distribution:

to me it looks better than a current mining situation that is rather bleak.