Bitfury Designs released under CC-BY-SA

[BitFury Group]

Hello.

BitfFury is deeply committed to a decentralized Bitcoin Blockchain ecosystem and we want to do all we can to ensure that as many hobbyists and engineers as possible have access to the technology we have developed with the support we've had from the Bitcoin community. Therefore we have decided to release our 28nm designs under Creative Commons Attribution Share Alike 3.0 License to give you the opportunity to get familiar with the style of our latest designs.

http://dl.bitfury.com/small/28nm/img/MbLiquid-9v41_thmb.jpg
http://dl.bitfury.com/small/28nm/img/48v_air_thmb.jpghttp://dl.bitfury.com/small/28nm/img/48V_hashboard_thmb.jpg

You will find higher resolution images together with schematics in http://dl.bitfury.com/28nm/.

Let's take a look at the files:

12Vcircuit.pdf contains a circuit of a 12V StringPower hashboard. PCIe connector is used to interface with the motherboard. The connector has modified SPI interface with 2 data buses leading to chips. On the right side of the connector is the Level_norm level shifter cirquit which is there mostly for 55nm legacy purposes.

48Vcircuit.pdf contains a circuit of a 48V StringPower hashboard. Not much difference to 12V except longer strings for higher voltage.

StringPowerBlock.pdf is probably the most interesting one. This is a block of one 250nm communicator chip and seven 28nm chips. Here you can see how the individual blocks are interconnected (differential pairs), hashing chip connections and how the StringPower design that BitFury pioneered with our first 55nm chip works.

It's worth pointing out that the decoupling capacitors that you see around the hash chips and the 250nm communicator chip will be absent in the 16nm design. They have all been integrated. The only thing that will be required on the hash board apart from 16nm and 250nm chips will be 10pF 0402 capacitors for communication pins. So no semiconductors, electrolytes, stabiliziators, resistors etc will be needed. How's that for minimal BOM cost?

MbLiquid-9_v1.pdf This is one of our latest motherboard designs, that can accomodate 9 hash boards. As you can see we're using the releatively cheap PCIe connectors for connecting the hash boards.

This design is built around the STM32F407 that is being used as a commutator with UART input from any source (RasPi, USB-UART converter etc). The output is 9 SPI buses for connection of the hash boards. So from 1 UART connection we're able to service up to 1700 hash chips (over 20TH/s with our 28nm chip and even more with the 16nm). 16nm will work in a very similar way.

All-in-all the design is quite flexible when it comes to voltage. Every block is fed with 4-6V and is connectedto the outside through differential connections through small 10pF capacitors. Blocks can be connected in parallel or in series on one or several PCBs. The can also be connected directly to SPI of e.g. Raspberry Pi. They don't require stable voltage so no stabilisators or regulators are needed. The only downside of this design are the 6-12 I/O connections that take up some space on the PCB.

[1713258119]

1713258119

[1713258119]

1713258119

[1713258119]

1713258119

[BitFury Group]

Reserved for future updates.

[BitFury Group]

And some more space reserved.

[philipma1957]

nice to something from your company.

you simply look to be far too late to bring the new gear to market.

last few jumps make your gear look poor roi to many of us.

next Jump we go to a diff of 170 and a Hash Rate of 1300 ph.

Difficulty History

Date                    Difficulty                   Change   Hash Rate
Feb 07 2016   144,116,447,847   20.06%   1,031,625,717 GH/s
Jan 26 2016   120,033,340,651   5.89%   859,232,121 GH/s
Jan 13 2016   113,354,299,801   9.12%   811,421,684 GH/s
Dec 31 2015   103,880,340,815   11.16%   743,604,444 GH/s
Dec 18 2015   93,448,670,796           18.14%   668,931,642 GH/s



If you want people to get your gear it needs to be here now.

Last time I purchased Bitfury gear I spent 17 BTC it was delayed and by the time I got it it was obsolete.

  Looks to me like you are bringing out your new gear far too slowly just like you did back in 2013.

I would have liked something maybe 5 to 6 thousand worth  maybe even 10 thousand usd , but It just seems to be I won't get it anywhere near on time.

Next jump a 4th unit fully built for 600 usd that would burn 280 watts looks like an okay piece of gear.

We both know there is no way you will be delivering that unit in 8 days.  More like 8 weeks.

Sad as the gear I purchased from your company (bitfury chips were in it)  was good gear  just came along too late.

[punin]

Sad as the gear I purchased from your company (bitfury chips were in it)  was good gear  just came along too late.

Please let us know who was the integrator that used our chips and delivered late? What website did you purchase from?

[Finksy]

Phil, I believe these pinouts and pictures were released in good faith and in anticipation for the upcoming 16nm chips to help developers familiarize themselves with the hardware and layout.  They have not (and I believe Mr. Punin said they have no intention still) of selling their 28nm gear as is pictured above.

Thank you for the info, very interesting.

[sobe-it]

Very interesting, kept the easy and simplicity of the older chip. Voltage around the edges, ground belly pad and a few comm pins.


What happens if a chip fails on the voltage side of the string? The older chip would fail and raise the voltage on the rest of the sting until they reached their melting point.

On the other hand I like this design, it seems that this way if a chip fails you only lose comm to only that chip. Unlike most miners if 1 chip drops you lose comm to the remaining chips in the chain, drastically rendering your miners hash power by 1/3 or 1/2.


I cant wait to see more chips/miners maybe even purchasing a miner down the road depending on what happens. I know I loved my older bitfury gear, quiet and simple not to mention if a board or chip failed you didn't lose a good amount of hashing power. I know a lot of guys would love to play and test with the new 16nm chip, I would too but hell I would still tinker with the 28nm one.

Sad as the gear I purchased from your company (bitfury chips were in it)  was good gear  just came along too late.

Please let us know who was the integrator that used our chips and delivered late? What website did you purchase from?

I purchased a few 55nm rev 2 chips from bitfurystrikesback when they were hot and they got shipped to someone 2 states over and they opened the package. By the time they returned it to the shipper and I received it, the chips were out of their packaging loose in the cardboard package and I was never able to use them, I sent them an email they replied back IT WOULD BE OK........ no chips or btc, not even a real sentence to my issue.

[2112]

The 4 pdfs posted thus far are of nearly zero educational value. They are just top level board schematics without the description of the pin functions of Bitfury's proprietary chips. Without that information they may be of some value to a reverse-engineer, but as a general technical information source their value is zero.

The one-line bitfury post from nearly 3 years ago contains more technical value, because it shows the internal topology of the input and output stages of the still secret original bitfury chips.

There was more information posted in 2013, but is all inaccessible because at that time bitfury used mega.co.nz file locker which is now gone.

So the most of the educational value you get from the above post is:

1) Bitfury uses separate chips of their own design made in 250nm process for signal distribution and gathering (thicker oxide -> higher supply voltage)

2) Any competitor who still uses high-current buck converters with on-board magnetics was and will be significantly undercut by Bitfury purely on the cost of 3rd party components.

That's about it.

I welcome any comments from anyone who may have recognized the circuit fragments using discrete BJTs, MOSFETs and OpAmps and could post some links to the full circuit schematics including the on-chip interface stages.

Пpикoльнoe питaлoвo, нaдeюcь в нaшeй дepeвнe мoжнo нaйти. Coфт для интepфeйca aдaптиpoвaл?

B нaшeй дepeвнe этoгo нe нaйти)))) Я для дeлa pacпaяю ModMiner Qad (oт BTCFPGA) тaм ecть и OKR-t/3 и D12F200A
A пo coфтy для MCP2210? я eё нe нaшёл (кoгдa был в Mocквe зaeзжaл нa Mитинo, нeт) нy мoжeт oнo и к лyчшeмy, зaймycь линeйкoй

O! Boт этo (линeйкa) yжe интepecнo!!!! Кaк-жe oнa бyдeт жить! He пepeбapщивaйтe c нaпpяжeниeм для нaчaлa.

[notlist3d]

Thanks for sharing some designs, this actually makes me more hopeful of home/hobby miners with Bitfury gear possibly.  I hope that the developers get support in getting chips.

I would love to see some miners for home/hobby miners.  Guess we will see but this is a positive direction.

[kilo17]

I do applaud Bitfury for making a presence on the forum, I hope it is because they are truly committed to us "little guys". 

[philipma1957]

Phil, I believe these pinouts and pictures were released in good faith and in anticipation for the upcoming 16nm chips to help developers familiarize themselves with the hardware and layout.  They have not (and I believe Mr. Punin said they have no intention still) of selling their 28nm gear as is pictured above.

Thank you for the info, very interesting.

I know they are going to be selling the newer gear the 16nm chips.

I am not angry at them.

I am stating this because I feel a huge amount of hash is going online in China as I type.  I feel it is s-7 like .25 - .30 watt gear.

Built cheaply and installed by the builders of it on a 1  cent power deal as a preemptive move against bitfury's .1 watt chip.

If I have 100 mega watts at 1 cent my 300 dollar s-7 beats the bitfury chip.

I think this is what the diff jumps  are about.    So I am merely stating that a working bitfury 16nm needs to be shipping no later then March 1 to have any chance of making money.

We go to 170 diff in 8 days.
We go to 205 diff in 21 days.

It  looks like China will raise the hash rate so high it makes the bitfury chip effectively outdated. For all but a 3 cent power user.

All bitmaintech need is 100 mega watts at 1 cent that = 300 to 400 ph.

https://en.wikipedia.org/wiki/Three_Gorges_Dam   this plant can do 22,500 mega watts


it looks like we are facing this right here right now.  

I do not push Fud  but if China asic builders cut a deal with the power plant above from now until the ½ ing  the diff could go to 300 really fast.

So I brought up that this 16mn gear needs to be on a site for sale with quick shipping.  Cause in two months at  a diff of 300 we don't need it.


remember what happened to the sp-20 when it tried to compete with the s-5 bitmaintech crushed them.  this looks like bitmaintech trying to crush bitfury.

[NotFuzzyWarm]

I like the 48vdc string feed idea. Considering that is very common telco power it makes for using fairly low(ish) cost but utterly reliable POL bricks to drop higher voltage ac/dc feed lines to board usage levels

[philipma1957]

I like the 48vdc string feed idea. Considering that is very common telco power it makes for using fairly low(ish) cost but utterly reliable POL bricks to drop higher voltage ac/dc feed lines to board usage levels

Yeah 48dc is smart idea.  Plenty of good efficient 48 watt  psu from mean well for instance

[NotFuzzyWarm]

Fair point on Cheap hyrdo power in China eg the 3 Gorges Dam but frankly, that vast majority of overall power production production is spoken for. For most heavily industrialized countries, for just one segment - aluminum production - takes one helluva chunk. ref: http://wordpress.mrreid.org/2011/07/15/electricity-consumption-in-the-production-of-aluminium/
"According to Alcoa, the world’s largest producer of aluminium, the best smelters use about 13 kilowatt hours (46.8 megajoules) of electrical energy to produce one kilogram of aluminium; the worldwide average is closer to 15 kWh/kg (54 MJ/kg).

Worldwide production of aluminium in 2010 was 41.4 million tonnes. Using the figures above this means that 621 billion kilowatt hours of electrical energy were used in the production of aluminium. To put that in perspective, the total world production of electrical energy was 20261 billion kilowatt hours, meaning that more than 3% of the world’s entire electrical supply went to extraction of aluminium."

Toss in the other more pressing power needs of China and you get my point.

Ja certainly a lot of low cost power will be available to local (ish) peta-farms but not as much as you think. Outside of being right next to the power plant/dam (the now closed Alcoa smelter site in Washington state comes to mind) the infra over there just does not support someone suddenly deciding to run 25-50MW much less 100MW or more without a helluva lotta prep and time/cash for it to be done.

[2112]

I like the 48vdc string feed idea. Considering that is very common telco power it makes for using fairly low(ish) cost but utterly reliable POL bricks to drop higher voltage ac/dc feed lines to board usage levels

Yeah 48dc is smart idea.  Plenty of good efficient 48 watt  psu from mean well for instance

I think you are both missing the point of the 48V DC supply. Somewhere between 25V and 50V there's a point where power no longer needs to be regulated. It is sufficient to just rectify and ripple filter it. With lower voltages you cannot risk it because accidental overvoltage will permanently destroy the oxide layer on the chip. With series/string implementation backed by some sort active voltage divide balancing you get enough oxide layers in series to be no longer afraid of surges. There's enough margin between the normal operating point and the breakdown voltage.

Nobody cares for under-voltages or sags. They just cause momentary increase of erroneous results.

The only thing that requires regulated power supply (and uninterruptible power supply) is the mining controller. But it has negligible power requirements compared with the hashing engines.

Remember that coin mining equipment isn't really a computing equipment or telecommunication equipment. You can reset it as often as you like and you never store any information for more than milliseconds.

[NotFuzzyWarm]

I like the 48vdc string feed idea. Considering that is very common telco power it makes for using fairly low(ish) cost but utterly reliable POL bricks to drop higher voltage ac/dc feed lines to board usage levels

Yeah 48dc is smart idea.  Plenty of good efficient 48 watt  psu from mean well for instance

I was thinking more along the line of TDk's  PFE500S-48  http://www.us.tdk-lambda.com/ftp/specs/pfe.pdf good for 48vdc @ 504w each. Given multiple power feeds to the strings, use as many as needed per board. They'd love being in Novec next to the hash boards...

Power bricks like that have been used for telco and such for decades (can you say easy and near-instant battery backup protection?) so there should be a shitload of them available as pulls not to mention new.

[2112]

I was thinking more along the line of TDk's  PFE500S-48  http://www.us.tdk-lambda.com/ftp/specs/pfe.pdf good for 48vdc @ 504w each. Given multiple power feeds to the strings, use as many as needed per board. They'd love being in Novec next to the hash boards...

Power bricks like that have been used for telco and such for decades (can you say easy and near-instant battery backup protection?) so there should be a shitload of them available as pulls not to mention new.

It sure would work, but it is an overkill.

The ultimate mining power supply is essentially a welding rectifier with some more ripple filtering and arc-start disabled. The normal welding rectifier have some very coarse regulation operating at line frequency (50Hz or 60Hz), not in the kHz or MHz ranges of precision voltage regulators used for non-power electronics.

With such power supplies your at-the-wall GH/J will be the same as at-the-chip GH/J within the normal measurement and process tolerances.

If any home experimenter reads this message: I'll reiterate the necessity of disabling the arc-start circuitry of the normal welding rectifiers. They intentionally produce higher voltage when the output current is near zero, then rapidly drop it to the nominal output voltage once the electric arc starts.

[NotFuzzyWarm]

I like the 48vdc string feed idea. Considering that is very common telco power it makes for using fairly low(ish) cost but utterly reliable POL bricks to drop higher voltage ac/dc feed lines to board usage levels

Yeah 48dc is smart idea.  Plenty of good efficient 48 watt  psu from mean well for instance

I think you are both missing the point of the 48V DC supply. Somewhere between 25V and 50V there's a point where power no longer needs to be regulated. It is sufficient to just rectify and ripple filter it. With lower voltages you cannot risk it because accidental overvoltage will permanently destroy the oxide layer on the chip. With series/string implementation backed by some sort active voltage divide balancing you get enough oxide layers in series to be no longer afraid of surges. There's enough margin between the normal operating point and the breakdown voltage.

Nobody cares for under-voltages or sags. They just cause momentary increase of erroneous results.

The only thing that requires regulated power supply (and uninterruptible power supply) is the mining controller. But it has negligible power requirements compared with the hashing engines.

Remember that coin mining equipment isn't really a computing equipment or telecommunication equipment. You can reset it as often as you like and you never store any information for more than milliseconds.

No I got that part of the equation. Look into the LED strings inside of any high-power lighting LED chip (clear lens, not ones with phosphors, can't see through it) from Cree or Phillips and you see my work there which hit the factory floor in 2007 giving the ability for fit (currently) just over a dozen emitter dies pushing over 25W total into a chip package >2mm². That in turn with the already known to begin with advantages of any series circuit and the major power LED makers were in heaven to start pushing the lighting biz to new limits.

Got am inexpensive LED bulb from Cree, Philips/LumiLEDs, Osram and no doubt others or one of the great blindingly bright LED flashlights boasting a 1-5w Cree emitter in it? You're welcome. Yes other tech advances were involved with gazillions of Other Peoples Money over the years leading to them but my bit in the chip package mfg'ing equation opened the door for it all to come together relatively inexpensively..

Anywho, back to the topic at hand, and question for me just brings up, how they protect against chip failures?
High current Zeners in the chip package (not on die)?

Since as you've said, the actual miner logic needs are pretty damn small, that leaves the door open for more to be happening in the chip package but off the (now 16nm node) ASIC die real estate per-se. Say just using the die real estate to also hold a safe operation fault logic controller for the off-die fault switches. Perhaps active internal power MOSFET switching to take chips out of the string or other failure protection mechanism?

Ease of very low tech battery backup until the diesels kick in is just icing on the cake for this app. Very easy availability of the power bricks is another.

[NotFuzzyWarm]

I was thinking more along the line of TDk's  PFE500S-48  http://www.us.tdk-lambda.com/ftp/specs/pfe.pdf good for 48vdc @ 504w each. Given multiple power feeds to the strings, use as many as needed per board. They'd love being in Novec next to the hash boards...

Power bricks like that have been used for telco and such for decades (can you say easy and near-instant battery backup protection?) so there should be a shitload of them available as pulls not to mention new.

It sure would work, but it is an overkill.

The ultimate mining power supply is essentially a welding rectifier with some more ripple filtering and arc-start disabled. The normal welding rectifier have some very coarse regulation operating at line frequency (50Hz or 60Hz), not in the kHz or MHz ranges of precision voltage regulators used for non-power electronics.

With such power supplies your at-the-wall GH/J will be the same as at-the-chip GH/J within the normal measurement and process tolerances.

If any home experimenter reads this message: I'll reiterate the necessity of disabling the arc-start circuitry of the normal welding rectifiers. They intentionally produce higher voltage when the output current is near zero, then rapidly drop it to the nominal output voltage once the electric arc starts.

Definitely ja. For home users battery chargers not so great either on single phase. 3-ph to DC and things get pretty rosey real fast with ripple dropping like a stone depending on the secondary side configuration and very stable. A delta-wye bridge on 3-phase 60hz give 720Hz ripple at ~ 0.5% pk-pk with no filtering whatsoever. This has not gone unnoticed by data center power providers pushing the high-DC feed to racks solutions.

[QuintLeo]

All bitmaintech need is 100 mega watts at 1 cent that = 300 to 400 ph.

https://en.wikipedia.org/wiki/Three_Gorges_Dam   this plant can do 22,500 mega watts

 The combination of dams along the Columbia River owned by the Chelan/Douglass/Grant PUCs have a comparable combined capasity.
 IIRC it's about 2500 MW per dam and 11 or 12 total dams.

 https://en.wikipedia.org/wiki/List_of_dams_in_the_Columbia_River_watershed

 36000 MW total capasity, but I think part of that is outside those 3 counties.

 I'm also pretty sure that the available power in that area was part of the reason the Hanford reservation was located near there (they did a LOT of power-intensive nuclear fuel enhancement work at that site over a long period of time).


 I dunno what the combined capasity of the TVA dams would be, or how many are still in operation, but I don't think they have anywhere near the installed capasity - there's a good reason Oak Ridge enhancement work got moved to Hanford - not enough power available (along with greater worries about security due to higher population density).


 One BIG damn = one BIG single point of failure, though it *might* be more economical to operate a single dam if all other factors are equal (in this case, they're not - labor costs in China are a lot less than anywhere in the US on average).


 And yes, Alcoa has at least one aluminum plant in the CDG area - Chelan county semi-near Wenatchee IIRC - unless that's the one that recently closed.