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

[Carlton Banks]

i am still betting someone, somewhere will release in 20 or 16nm during 2014 !

Not unless you have enough money to build their own fabrication plant  

Any new node will be booked solid for years by the likes of IBM, Intel, AMD, nVidia, Samsung etc. for their own products.

http://en.wikipedia.org/wiki/List_of_semiconductor_fabrication_plants

thats incorrect.  tsmc has said they've done tens of customer tapeouts in 20 and 16nm already!

in several articles, tsmc has said more than 25 customer tape-outs so far and more planned for 2014...

http://wccftech.com/tsmc-begins-volume-production-20nm-chips-q1-2014-16nm-finfet-chips-q1-2015/

http://www.xbitlabs.com/news/other/display/20131022230815_TSMC_Shares_More_Details_Regarding_16nm_FinFET_and_20nm_Progress.html

I've also seen that Global Foundries are talking up their snatching the debut of 16nm process node out from under Intel, which could just be a bit of marketing gamesmanship, but could also be a genuine possibility.

It's getting to the point where the leading edge silicon feature shrinkages are converging with new SHA-256 ASIC chips, after all, the newest node process is usually tested out with known simple chip designs with plenty of recurring features (flash chips are a good example for those that look at the new consumer SSD products, 20nm has been available for some time now, and 25nm has been available for several years already). So 20nm is arguably tried and tested, and 28 nm is arguably getting a little long in the tooth. 16nm is the new frontier, bleeding edge process.

And all the comparisons with the process used by latest GPU's aren't really valid, the new GPU may well use new chip designs also, and the chip design itself just does not include as many redundant features as an ASIC or a flash memory chip.

[aerobatic]

i am still betting someone, somewhere will release in 20 or 16nm during 2014 !

Not unless you have enough money to build their own fabrication plant  

Any new node will be booked solid for years by the likes of IBM, Intel, AMD, nVidia, Samsung etc. for their own products.

http://en.wikipedia.org/wiki/List_of_semiconductor_fabrication_plants

thats incorrect.  tsmc has said they've done tens of customer tapeouts in 20 and 16nm already!

in several articles, tsmc has said more than 25 customer tape-outs so far and more planned for 2014...

http://wccftech.com/tsmc-begins-volume-production-20nm-chips-q1-2014-16nm-finfet-chips-q1-2015/

http://www.xbitlabs.com/news/other/display/20131022230815_TSMC_Shares_More_Details_Regarding_16nm_FinFET_and_20nm_Progress.html

Hearing rumours they're about to go live and that their setup looks very similar to (but not the same as) asicminer in chips and cooling tech.

[analytics]

Big spike here early January.
http://bitcoin.sipa.be/speed-lin-10k.png

Maybe from them, who knows so far

[Cheshyr]

If the money came in last March or April, and they had a design already mostly RTL'ed, I would expect their engineering samples would start to trickle in this Feb or March.  Perhaps products in May or June.  They're still someone I want to watch out for, but they're doing a good job at running a tight ship, so I guess we'll just have to wait and see.

[volosator]

There was a post inside this thread saying about 4.7 mil NRE. Also, there was a link to .doc file explaining all that.
It disappeared this morning. Can anybody point me to that?

[Carlton Banks]

If the money came in last March or April, and they had a design already mostly RTL'ed, I would expect their engineering samples would start to trickle in this Feb or March.  Perhaps products in May or June.  They're still someone I want to watch out for, but they're doing a good job at running a tight ship, so I guess we'll just have to wait and see.

Except they're said to be all in-house, no consumer products. Although it could just be a load of marketing hype for future salesof consumer products, I guess ("We now offer you the chance to buy the most closely guarded mining technology the world has ever known..." etc)

[Cheshyr]

If the money came in last March or April, and they had a design already mostly RTL'ed, I would expect their engineering samples would start to trickle in this Feb or March.  Perhaps products in May or June.  They're still someone I want to watch out for, but they're doing a good job at running a tight ship, so I guess we'll just have to wait and see.

Except they're said to be all in-house, no consumer products. Although it could just be a load of marketing hype for future salesof consumer products, I guess ("We now offer you the chance to buy the most closely guarded mining technology the world has ever known..." etc)

A good point.  I was just thinking in terms of ASIC delivery schedules based on the reported 4.7M capital.  If they stay private, I doubt we'll notice, since we've got HashFast, Cointerra, BitMine, and Bitmain all dumping piles of power on the network over the next few months.  An interesting mystery, but I'm still not certain what to do with any information we find about them.

[descarte]

doesn't mean all bright people come together can produce anything useful. There is still the EQ aspect that people don't see. They will face the same stiff competition like other companies. It is a mountain to climb.

[Dalkore]

I just read this:

21e6

I've mentioned them a couple of times...  no one seemed to comment.

https://bitcointalk.org/index.php?topic=269093.msg3059525;topicseen#msg3059525

Well tell us more.

if i knew more i would've told more !

all i know is that the founder is a stanford professor with a good reputation - BALAJI SRINIVASAN... and that there's serious money involved.  we know they raised $5m.. but maybe more...

and that they've got a lot of bright people involved.

and that they've been working on it longer than several of the other guys.. so, one would assume that they're keeping it secret.  this is the only 'secret mine' that I've ever heard of.

the big question is... what does $5m buy them?   does it get them enough to design and build their own silicon?  and what silicon will it be?  is it state of the art 28nm, or is it a cheaper, older process node?

heck, with their connections, they could even be trying to go for 20 or 16nm...!?

we don't seem to see any of their miners hashing so presumably they don't have silicon back yet, or maybe it didn't work, or maybe its not good enough.. which is odd, as people who started later than them have already received their silicon or are expecting it within a month or so.

maybe it went wrong?  maybe they got delayed?  its a mystery that such an experienced and well funded team seems to have not (yet) achieved their goals.

I think 20 or 16nm is way out of discussion. $5m should be enough for a 28nm mask(design included). Where or when did you heard about them and how? Do you know for how long are they working on this project?

$5 million should be enough for them to develop a chip and product.   

[aerobatic]

I just read this:

21e6

I've mentioned them a couple of times...  no one seemed to comment.

https://bitcointalk.org/index.php?topic=269093.msg3059525;topicseen#msg3059525

Well tell us more.

if i knew more i would've told more !

all i know is that the founder is a stanford professor with a good reputation - BALAJI SRINIVASAN... and that there's serious money involved.  we know they raised $5m.. but maybe more...

and that they've got a lot of bright people involved.

and that they've been working on it longer than several of the other guys.. so, one would assume that they're keeping it secret.  this is the only 'secret mine' that I've ever heard of.

the big question is... what does $5m buy them?   does it get them enough to design and build their own silicon?  and what silicon will it be?  is it state of the art 28nm, or is it a cheaper, older process node?

heck, with their connections, they could even be trying to go for 20 or 16nm...!?

we don't seem to see any of their miners hashing so presumably they don't have silicon back yet, or maybe it didn't work, or maybe its not good enough.. which is odd, as people who started later than them have already received their silicon or are expecting it within a month or so.

maybe it went wrong?  maybe they got delayed?  its a mystery that such an experienced and well funded team seems to have not (yet) achieved their goals.

I think 20 or 16nm is way out of discussion. $5m should be enough for a 28nm mask(design included). Where or when did you heard about them and how? Do you know for how long are they working on this project?

$5 million should be enough for them to develop a chip and product.   

theres a rumour they raised a lot more recently.  anyone know for sure?

[Sine(X)]

Take my rumor for what it is. I take what I've heard in earnest, but you know, grain o' salt and that. The points I'd like to point out are...

1. 21^6 controls 12% of the bitcoin mining production. Sounds like a sizeable chunk to me. Evidently they were going for 20%. 
2. Belaji is a fucking evil genius, in the best kinda way. He's fucking great. And not evil. Nevermind.
3. The company intends to remain private, and distribute the bitcoin dividends to their shareowners, permanently. It's pretty cushy if you're a shareholder.

I figure this thread to be dead, but I happened upon a very unique bit of information y'all might be interested in. Figure what you will with it. It is recent information.

Any news? How to become a shareholder?

[Dr Charles]

very interesting. I am going to do some additional digging myself. thanks.

[dogie]

very interesting. I am going to do some additional digging myself. thanks.

Let me know if you find anything please. I think it would be strange that they have / held 20% of the network if that was their only funding round.

[Dr Charles]

^^ Agreed. Ill let you know if I find anything worth while.

[antirack]

21E6 operated since 2014 in a commercial wholesale DC, using modular turnkey infrastructure (not the new DataTank systems).

Percentage of network in low single digit during 2014 decreasing as difficulty went up (to my best knowledge based on MW numbers of a facility on the east coast).

[Sine(X)]

21E6 operated since 2014 in a commercial wholesale DC, using modular turnkey infrastructure (not the new DataTank systems).
Percentage of network in low single digit during 2014 decreasing as difficulty went up (to my best knowledge based on MW numbers of a facility on the east coast).

Did you mean this?
16 megawatt lease from CyrusOne in Phoenix by a bitcoin mining firm whose name has not been disclosed

[antirack]

21E6 operated since 2014 in a commercial wholesale DC, using modular turnkey infrastructure (not the new DataTank systems).
Percentage of network in low single digit during 2014 decreasing as difficulty went up (to my best knowledge based on MW numbers of a facility on the east coast).

Did you mean this?
16 megawatt lease from CyrusOne in Phoenix by a bitcoin mining firm whose name has not been disclosed

No, that's not the one I meant (Phoenix AZ is not at the east coast either).

[topminingcontracts]

Did anyone has info about new funds 21e6 raised to expand the mining operation?


Juan

[dogie]

Did anyone has info about new funds 21e6 raised to expand the mining operation?

Not from my end, no. They seem to have their heads down.

[Gleb Gamow]

This dude should know more: https://twitter.com/cetinkoc