[ANN] Bitfury ASIC sales in EU and Europe

[integrity42]

The bad:
Our preliminary tests show that the boards are falling a bit short on hashrate. This might be due to differences between wafers, immature software or SPI issues. Because we are in a hurry to ship, you win: I will ship your ordered hashrate regardless (ie. more hardware free of charge) until we fix this issue and can provide 400GH in one unit.

Hi punin,

I was wondering if you can tell us what amount of hardware comprises the 'unit' you speak of above? (i.e. how many M-boards, H-cards, etc)

Thanks. 

[-Redacted-]

Since they are now providing "2 units of 200 Gh/s minimum" to make up for the fact they cannot provide a single unit of 400 Gh/s, it is an m-board with one Raspberry Pi and one or more hashing cards inserted.   That makes a starter kit a unit with 25 Gh/s in it.  What, you trying to use semantics to weasel out of a bet or something?

[integrity42]

If you're talking about the BitBet bit, that has been decided already.

My question is pretty clear.

Hi punin,
I was wondering if you can tell us what amount of hardware comprises the 'unit' you speak of above? (i.e. how many M-boards, H-cards, etc)
Thanks.

Punin is talking about a 400Gh unit. I'm assuming the unit he's referring to contains 1M-board, 16H-cards, and 1R-pi.  But just want to confirm.

[MXRider]

Since they are now providing "2 units of 200 Gh/s minimum" to make up for the fact they cannot provide a single unit of 400 Gh/s, it is an m-board with one Raspberry Pi and one or more hashing cards inserted.   That makes a starter kit a unit with 25 Gh/s in it.  What, you trying to use semantics to weasel out of a bet or something?

August orders were built from "bad" chips and that's why some of the H-cards have difficulties hitting 25 GH/s. Full kits is around 370 GH/s depending on the quality of your H-cards. They are not offering "2 units of 200 GH/s minimum".

Full kit = one unit = M-board + 16 H-cards + RPi.

October H-cards have better chips and they will achieve 25 GH/s without any problems. There is actually a new revision coming but let's wait punin 's comments on that one.

[-Redacted-]

They are in the US.  And rather than shipping two cards for starter kits like punin did, they're sorting the cards and sending those that reach 25 Gh/s with starters.  For full 400 Gh/s klts in the US they're sending two units with minimum 200 Gh/s each.  Not sure what punin did for full kits in the EU, but I know good and well that he didn't ship out 370 Gh/s and call it 400...

[dani]

Hi All, Isokivi and I managed to hack together a script for logging data from the chainminer output. This might well be useful for those tuning chips by hand or for those that just like to gather data and make pretty graphs in Excel .

+1. I managed to get it running - i guess. reporting back when it's finished (set total runtime to 180min, don't expect a reply shortly).

Great stuff, would be interested to hear how it goes. We can tweak if necessary.

I started version 1.3, but I didnt had the correct chmod rights given (sorry, no native linux user here   ), but Isokivis post was easy to follow, did that and it worked. I wonder if you could add the speeds that autotune sets as well, like an average speed, but thats just nice to have.

let's share some data!

Code:

Chip #          Settings        avg. noncerate  avg. hashrate   Nonces/round    Falses/round    Falses %        SPI errors      MISO errors     
18              AIfDSo          0.765           1.442           373             303             44.82%          0               0               
19              AIfDSo          1.648           1.821           799             102             11.32%          1               0               
20              AIfDSo          1.954           1.994           946             3               0.32%           0               0               
21              AIfDSo          1.737           1.786           842             7               0.82%           0               0               
22              AIfDSo          1.64            1.788           793             3               0.38%           0               0               
23              AIfDSo          1.877           1.95            910             6               0.66%           0               0               
24              AIfDSo          1.989           1.96            964             6               0.62%           0               0               
25              AIfDSo          1.655           1.676           801             5               0.62%           0               0               
26              AIfDSo          1.535           1.629           743             22              2.88%           0               0               
27              AIfDSo          1.659           1.697           803             0               0%              0               0               
28              AIfDSo          1.247           2.079           606             420             40.94%          1               0               
29              AIfDSo          1.852           2.051           898             93              9.38%           0               0               
30              AIfDSo          0.109           0.305           53              41              43.62%          0               0               
31              AIfDSo          1.927           1.903           933             2               0.21%           0               0               
32              AIfDSo          2.014           2.037           977             19              1.91%           0               0               
17              AIfDSo          1.928           1.963           799             3               0.37%           0               0    

collected over 30 minutes

Code:

Chip #          Settings        avg. noncerate  avg. hashrate   Nonces/round    Falses/round    Falses %        SPI errors      MISO errors     
2               aifdso          0               0               0               0               0%              0               0               
3               AIfDSo          1.449           2.108           297             109             26.85%          0               0               
4               AIfDSo          0.394           1.277           80              112             58.33%          0               0               
5               AIfDSo          1.873           1.858           384             8               2.04%           0               0               
6               AIfDSo          1.958           1.984           401             1               0.25%           0               6               
7               AIfDSo          1.592           1.632           326             7               2.1%            0               4               
8               AIfDSo          1.915           1.721           392             0               0%              0               2               
9               AIfDSo          1.443           1.703           295             2               0.67%           0               0               
10              AIfDSo          1.485           1.441           303             3               0.98%           1               0               
11              AIfDSo          1.744           1.978           357             32              8.23%           0               0               
12              AIfDSo          0.322           0.35            66              3               4.35%           0               0               
13              AIfDSo          1.607           1.847           327             69              17.42%          2               6               
14              Aifdso          1.172           1.256           237             2               0.84%           0               0               
15              aifdso          0               0               0               0               0%              0               0               
16              AIfDSo          0.964           0.918           128             4               3.03%           0               2               
1               AIfDSo          2.123           2.087           286             10              3.38%           0               2      

note: the first chip is on the last spot (in every log created). Anyway, thank you for that fine piece of code

[jlsminingcorp]

I started version 1.3, but I didnt had the correct chmod rights given (sorry, no native linux user here   ), but Isokivis post was easy to follow, did that and it worked. I wonder if you could add the speeds that autotune sets as well, like an average speed, but thats just nice to have.

let's share some data!
note: the first chip is on the last spot (in every log created). Anyway, thank you for that fine piece of code

Great stuff, thanks for reporting back - sorry for the total lack of instructions, thankfully Isokivi was on hand to clear things up . Strange that the first chip ends up last in this analysis, I'll have to take a look at that tomorrow. Would you mind sending me a sample of one of the chip output files (say the first two time points) by pm when you get a chance (I don't want to spam this thread with debugging).

In terms of the average autotune speeds, it's over to Anduck for that one - he's the guy behind the web app. The data is logged in the text files that are outputted by the script running locally on your miner (the only thing that's excluded is the string of single-digit numbers at the end of each chip output in .stat.log - because, in my ignorance, I have no idea what they mean).

[jspielberg]

Hey Guys,

I am an american who bought a starter kit from megabigpower.  To get my 25 GH, I received a card that is functioning at full capacity (~19GH) and an "EOL" end of line card with is running about 6 to 10 GH.  After looking at the stats.log, it appears that the EOL card is only seeing the first 6 chips.

I took some macro hi-res pictures of the chips 6 and 7 (45 and 46 labelled on the card), where the spi communication I am guessing is broken.

https://i.imgur.com/Iddp2WQ.jpg


https://i.imgur.com/MpXXHdS.jpg


For the life of me, I don't see anything out of the ordinary.

The european thread seems more advanced at getting full speed out of these boards and was wondering if any experienced eyes see something suspicious.

[jimrome]

Hey Guys,

I am an american who bought a starter kit from megabigpower.  To get my 25 GH, I received a card that is functioning at full capacity (~19GH) and an "EOL" end of line card with is running about 6 to 10 GH.  After looking at the stats.log, it appears that the EOL card is only seeing the first 6 chips.

I took some macro hi-res pictures of the chips 6 and 7 (45 and 46 labelled on the card), where the spi communication I am guessing is broken.

For the life of me, I don't see anything out of the ordinary.

The european thread seems more advanced at getting full speed out of these boards and was wondering if any experienced eyes see something suspicious.

A few things to try:
- reflow solder joints, especially on SPI/CLK pins of chip 6 and 7
- failing that, short jumpers to bypass chip
- if it still doesn't work you may need to also either remove the chip, or if you don't have the tools to do this you could cut the traces

[joris]

In 20 minutes logging, I get MISO errors on 50% of the chips on my 2 boards. 2 up to 48 per chip (don't know how to paste the log properly here).

"MISOerr | Master In Slave Out error. Error in communication from chip to RasPi."

What causes these errors?

[goxed]

Hey Guys,

I am an american who bought a starter kit from megabigpower.  To get my 25 GH, I received a card that is functioning at full capacity (~19GH) and an "EOL" end of line card with is running about 6 to 10 GH.  After looking at the stats.log, it appears that the EOL card is only seeing the first 6 chips.

I took some macro hi-res pictures of the chips 6 and 7 (45 and 46 labelled on the card), where the spi communication I am guessing is broken.

https://i.imgur.com/Iddp2WQ.jpg


https://i.imgur.com/MpXXHdS.jpg


For the life of me, I don't see anything out of the ordinary.

The european thread seems more advanced at getting full speed out of these boards and was wondering if any experienced eyes see something suspicious.

Really good macro capture there. I had to cut traces, when soldering the jumpers did not work.

[klondike_bar]

Hey Guys,

I am an american who bought a starter kit from megabigpower.  To get my 25 GH, I received a card that is functioning at full capacity (~19GH) and an "EOL" end of line card with is running about 6 to 10 GH.  After looking at the stats.log, it appears that the EOL card is only seeing the first 6 chips.

I took some macro hi-res pictures of the chips 6 and 7 (45 and 46 labelled on the card), where the spi communication I am guessing is broken.

https://i.imgur.com/Iddp2WQ.jpg


https://i.imgur.com/MpXXHdS.jpg


For the life of me, I don't see anything out of the ordinary.

The european thread seems more advanced at getting full speed out of these boards and was wondering if any experienced eyes see something suspicious.

Really good macro capture there. I had to cut traces, when soldering the jumpers did not work.

What does cutting the trace do, simply cause it to skip it and recognise all the chips after it (7-16)?

with regards to the pictures, why are some of the traces' through-holes (right term?) gold/copper coloured? Is this just a result of the way the green colour/layer was applied overtop, or a potential defect? Seems like an unlikely reason, but it sounded like the board production was the issue behind most errors, rather than chips)

[-Redacted-]

If one of the chips has some problem (say, an internal short) causing it to pull a critical line down to ground, or up to Vcc, then just soldering the jumpers closed won't help.  That's why you might need to cut traces or unsolder the chip.  

If the chip is just completely dead or otherwise non-functional, then soldering across the jumpers could allow the other chips in the chain to work properly.  

The jumpers bypass the chip in question by taking the output of the previous chip and connecting them to the inputs of the next chip.

Nice photos, by the way - that's like professional quality photography there...

[jimrome]

What does cutting the trace do, simply cause it to skip it and recognise all the chips after it (7-16)?

Physically breaks the connection between the chip and the CLK/SPI bus. Even if the chip isn't hashing it could cause issues with the function of either.

[jspielberg]

Thanks guys for the advice.

As stated in the U.S. thread, I am going to try to reflow these chips first, and see if that gets me over the hump.

Anyone know the flow temperature of the solder used on the H-Boards? (Probably a question for the "free money" thread).

[goxed]

Thanks guys for the advice.

As stated in the U.S. thread, I am going to try to reflow these chips first, and see if that gets me over the hump.

Anyone know the flow temperature of the solder used on the H-Boards? (Probably a question for the "free money" thread).

My advice is against reflowing if it passes visual tests

[jspielberg]

Thanks again for the advice guys.

I couldn't find any visual smoking guns with the photos. As was also suggested... bridging the SJ pads on chip 7 might be a good first step.

[jimrome]

Thanks guys for the advice.

As stated in the U.S. thread, I am going to try to reflow these chips first, and see if that gets me over the hump.

Anyone know the flow temperature of the solder used on the H-Boards? (Probably a question for the "free money" thread).

My advice is against reflowing if it passes visual tests

It's pretty tough to do a visual test on a colder solder joint at this pitch, even with a scope it's difficult. Easier just to reflow.

[intron]

with regards to the pictures, why are some of the traces' through-holes (right term?) gold/copper coloured? Is this just a result of the way the green colour/layer was applied overtop, or a potential defect? Seems like an unlikely reason, but it sounded like the board production was the issue behind most errors, rather than chips)

Yes, the green solder mask is applied
over the the bores of the vias (those
small 'circles'). Sometimes some of the
ink disappears in the drilled holes and a
discoloration occurs. This is not an
indication of a problem.

intron

[marto74]

Hi,
Niko
I've send you a few Emails asking about bulk chip order prices.
Is there any problem ?

Sorry Marto, I'm drowning in emails, just had someone take over the communication for me. We will introduce new chip pricing on the site this weekend.

Any news ?