Unofficial BITMINE CoinCraft series 28nm ASIC miners thread

[psycho-pass]

let e make the video first

----------------------

found this

https://bitmine.zendesk.com/access/unauthenticated

this os only their support system....

but if we somehow get login we might find out what bitmine is upto

[horrorhorst]

My miners arrived today 

2x 1th/s desk and 1x200gh/s desk running @ ~2.4 th/s (40-45°C,nominal)

[Wolke]

let e make the video first

----------------------

found this

https://bitmine.zendesk.com/access/unauthenticated

this os only their support system....

but if we somehow get login we might find out what bitmine is upto

omg -> i'm no lawyer -> but this idea is very very illegal !!!!!!!!!

[Wolke]

My miners arrived today  

2x 1th/s desk and 1x200gh/s desk running @ ~2.4 th/s (40-45°C,nominal)

realy nice :-)

whats your shipping date ?

[E]

Pics and video delayed: on the first box I'm getting a Bad File Descriptor message from cgminer when it tries to open the SPI channel to the board selector:

Code:

 [2014-02-21 15:02:41] Failed to init board selector                    
 [2014-02-21 15:02:41] SPI '/dev/spidev0.0': mode=1, bits=8, speed=8000000                    
 [2014-02-21 15:02:41] checking board 0...                    
 [2014-02-21 15:02:41] Failed to write to fdesc -1: Bad file descriptor

Both spidev device files exist and have the correct permissions...

I'm going to try the other box to see what's up before trying to track this down. In the meantime I'll pastebin the dmesg

http://pastebin.com/cYG64v2j

and the bitminer mineros boot script:

http://pastebin.com/jWvPLQJQ

[Gogreen]

One person sent an e-mail!
Need to fill out by hand and still need my signature!

REFUND REQUEST FORM https://docs.google.com/file/d/0B4izEkfhvfklWXJDWFVkbVlzMGs

Their direct link Refund Request Form:
https://bitmine.zendesk.com/attachments/token/zfiab2apjgpgpui/?name=Refund+Request+Form.pdf

They are clearing the Q why not wait a little bit  longer.

[Gogreen]

My miners arrived today 

2x 1th/s desk and 1x200gh/s desk running @ ~2.4 th/s (40-45°C,nominal)

Happi for you 

[Gogreen]

As I said, they're not runnig at my home. But I'll ask my friend for some screenshots...
But first he has some problems to solve...according to eligius.st the two miners deliver only 377 Gh/s. At ghash.io it was around 600 Gh/s....which is very strange. There are no HW errors and they should deliver at least 1200 Gh/s :/

CONGRATS  . Vidoe please  .  No rushing take your time,

[zefir]

Pics and video delayed: on the first box I'm getting a Bad File Descriptor message from cgminer when it tries to open the SPI channel to the board selector:

Code:

 [2014-02-21 15:02:41] Failed to init board selector                    
 [2014-02-21 15:02:41] SPI '/dev/spidev0.0': mode=1, bits=8, speed=8000000                    
 [2014-02-21 15:02:41] checking board 0...                    
 [2014-02-21 15:02:41] Failed to write to fdesc -1: Bad file descriptor

Is this reproducible? I saw this only once some while ago and was not able to track it down. If it shows up regularly, please let me know and it will give me a great chance to resolve the issue.


Thanks,
zefir

[E]

Pics and video delayed: on the first box I'm getting a Bad File Descriptor message from cgminer when it tries to open the SPI channel to the board selector:

Code:

 [2014-02-21 15:02:41] Failed to init board selector                    
 [2014-02-21 15:02:41] SPI '/dev/spidev0.0': mode=1, bits=8, speed=8000000                    
 [2014-02-21 15:02:41] checking board 0...                    
 [2014-02-21 15:02:41] Failed to write to fdesc -1: Bad file descriptor

Is this reproducible? I saw this only once some while ago and was not able to track it down. If it shows up regularly, please let me know and it will give me a great chance to resolve the issue.


Thanks,
zefir

It's not only reproducible, it's all I can get out of either box

What info can I give you to help?

[Rave]

6 modules in normal mode:

[giorgiotheclown]

Pics and video delayed: on the first box I'm getting a Bad File Descriptor message from cgminer when it tries to open the SPI channel to the board selector:

Code:

 [2014-02-21 15:02:41] Failed to init board selector                    
 [2014-02-21 15:02:41] SPI '/dev/spidev0.0': mode=1, bits=8, speed=8000000                    
 [2014-02-21 15:02:41] checking board 0...                    
 [2014-02-21 15:02:41] Failed to write to fdesc -1: Bad file descriptor

Is this reproducible? I saw this only once some while ago and was not able to track it down. If it shows up regularly, please let me know and it will give me a great chance to resolve the issue.


Thanks,
zefir

It's not only reproducible, it's all I can get out of either box

What info can I give you to help?

scary!

wish you good luck...

[SolarSilver]

Coincraft Rig: No, due to technical problems

Since they haven't posted I video of a working Rig, I really start to worry.
I don't think that the first rigs will be shipped before mid March. Or how do you see this?

I think those who ordered a rig are going to get screwed even more by extra delays. Bitmine should just man up and stop wasting time trying to fix a non issue: If you can't get 2 TH/s from one rackmount, just start shipping people 2 rackmounts that each do 1 TH/s

Nobody is going to complain about the 2 boxes, I don't know anybody who is going to be space restrained. Power and cooling restraint, yes.

But people will complain if they lose yet another week...

[E]

Pics and video delayed: on the first box I'm getting a Bad File Descriptor message from cgminer when it tries to open the SPI channel to the board selector:

Code:

 [2014-02-21 15:02:41] Failed to init board selector                    
 [2014-02-21 15:02:41] SPI '/dev/spidev0.0': mode=1, bits=8, speed=8000000                    
 [2014-02-21 15:02:41] checking board 0...                    
 [2014-02-21 15:02:41] Failed to write to fdesc -1: Bad file descriptor

Is this reproducible? I saw this only once some while ago and was not able to track it down. If it shows up regularly, please let me know and it will give me a great chance to resolve the issue.


Thanks,
zefir

It's not only reproducible, it's all I can get out of either box

What info can I give you to help?

Alright, good news for the 600GH box at least; I opened both boxes up and found the controller board -> module board (?) ribbon cable was disconnected. I'll include a picture of what to look for in my review. It's up and hashing @ 616GH/s indicated on the front panel; no pool stats yet.

Searching for a similar problem on the other machine.

Overall first impression: Very nicely done; quieter than Batch 2 Avalon, louder than KnC (20C ambient). There's plenty of room for cooling improvement, especially on units that aren't fully populated, so fingers crossed on overclocking.

One issue that harkens back to Avalon days -- the desk is cooled by two fans blowing out the back of the unit; this means negative pressure differential in the unit and _potential_ problems for PSU cooling near the limit due to the PSU fan fighting the pressure.

edit: The unit came preconfigured with the following mining setup:

http://eligius.st/~wizkid057/newstats/userstats.php/1BitmineL6VxA7tKJNkZxJPx3p1gwd6uDH

[Daveb0]

Nobody is going to complain about the 2 boxes, I don't know anybody who is going to be space restrained. Power and cooling restraint, yes.

Depends on whether you run your own "datacenter" or have the rigs hosted as colocation servers (or "housed" severs). There you have the oposite situation
There you pay also "per U" or 1/4 Rack and don't really care about power and cooling

But still I would prefer getting two boxes which reach the advertised turbo mode (once more: there's no "up to" for the rigs), than having one box just running at nominal speed.

edit: added a few words

[E]

Overclocking Info!

The webui offers three choices:

Powersave
Nominal
Turbo (BETA)

This is picked up in /mineros/boot.py:

Code:

def start_cgminer():
        conn = sqlite3.connect(MINEROS_DB_LOCATION)
        c = conn.cursor()
        c.execute('SELECT * FROM cgminer LIMIT 1')
        row = c.fetchone()

        a1_options='16000:800000:8000'

        if row[13]=='0':
                a1_options='16000:600000:4000'
        elif row[13]=='1':
                a1_options='16000:800000:8000'
        elif row[13]=='2':
                a1_options='16000:1000000:8000'

        os.system('start-stop-daemon --start --background -m --oknodo --pidfile /run/cgminer.pid --exec cgminer -- --text-only --lowmem --bitmine-a1-options '+a1_options+' -o '+row[1]+' -O '+row[2]+':'+row[3]+' -o '+row[4]+' -O '+row[5]+':'+row[6]+' -o '+row[7]+' -O '+row[8]+':'+row[9]+' '+row[10]+' --api-allow '+row[12]+' --api-listen --queue 50 '+row[13]+' &')

        c.close()

The a1_options become arguments to the --bitmine-a1-options flag, which is documented as such:

Code:

--bitmine-a1-options <arg> Bitmine A1 options ref_clk_khz:sys_clk_khz:spi_clk_khz:override_chip_num

I don't know what the clocking situation is on chip, not having read any docs, so I don't know what clock speeds are actually possible, but

Code:

cgminer --bitmine-a1-options 16000:850000:8000 --lowmem -o stratum+tcp://stratum.mining.eligius.st:3334

is working great for me, returning around 645GH/s with a hardware error rate of 0.004% (yes! - 8HW with >170000 A+R+HW).

I'm going for a walk to let stats stabilize; will try other clockrates later.

[E]

One issue that harkens back to Avalon days -- the desk is cooled by two fans blowing out the back of the unit; this means negative pressure differential in the unit and _potential_ problems for PSU cooling near the limit due to the PSU fan fighting the pressure.

I can confirm that airflow is reversed through the power supply under my present operating conditions - a flame held to the PS grill is drawn sharply inward. This may not bode well for the PS depending on how directionally dependent its cooling airflow needs are.

[3DX4D]

Coincraft Rig: No, due to technical problems

Since they haven't posted I video of a working Rig, I really start to worry.
I don't think that the first rigs will be shipped before mid March. Or how do you see this?

I think those who ordered a rig are going to get screwed even more by extra delays. Bitmine should just man up and stop wasting time trying to fix a non issue: If you can't get 2 TH/s from one rackmount, just start shipping people 2 rackmounts that each do 1 TH/s

Nobody is going to complain about the 2 boxes, I don't know anybody who is going to be space restrained. Power and cooling restraint, yes.

But people will complain if they lose yet another week...

This +infinity.  I have a rig on order (week 5 Jan delivery when I purchased). I only bought 1th but wanted room to expand if all worked out good. It sucks balls watching all these 1th desk units going out while the rig is still in fucking limbo, at this rate it won't be long until people who ordered desk units after me will receive a miner before me  .  I would gladly switch my 1th rig order to a 1th desk if the option was available.

[E]

I don't want to trigger too much consternation, but this may be my 5-module unit's problem:


(which is why they don't sell the Platimax 1500W in the US)

I'll be able to try this on a 240v circuit in a few hours...

[Tegija]

....

Thanks,
zefir

WOW! now we at least know that at least one of bitmine´s officials found this "unofficial BITMINE thread"!

Even nicer is that he is reading and answering here, so maybe its not too much to expect that we will get here some informations about Coincraft Rig´s status, Zefir?