[ANN] US/North American Bitfury sales NEW STOCK ***NOW SHIPPING***

[cypherdoc]

can someone explain the concept of the Banks and how they apply to v2.2 H boards and v3 M boards?

Dave said that the Banks are obsolete and that the H boards are just numbered 0-9, A-F.  however, on 3 of my M boards i have one entire bank at zero; 2 Bank 1's, and 1 Bank 4.  there must be some relationship btwn the 4 adjacent cards in those Banks.

any suggestions to revive those Banks?

Even v3 M-boards have 4 banks of 4 slots each.

Try reseating cards, or rotate the 4 cards within a bank.

If you think the problem is related to VRM thermal/current overload, try disabling a chip or two on each card, or turn down the voltage with the trimpots.

i have heatsinks and multiple fans so hopefully it isn't thermal.

i know the GUI shows 4 banks but do they have a function?  Dave did say they didn't mean anything.

Can you please post a screenshot of your bitfury webpage? Can you enable remote login to your rig for a few mins?

Since Bank1 and 2 are kaput, You will have to

a) Unplug all but first card from Bank 1 and Bank 2
b) Check if the cards on Bank 1 and  Bank 2 are mining
c) Add one more card to each Bank
d) Repeat b through c

what's the recommended procedure for doing this as in "stop miner" or "shut down completely" or "power off" before adding another card?

[goxed]

can someone explain the concept of the Banks and how they apply to v2.2 H boards and v3 M boards?

Dave said that the Banks are obsolete and that the H boards are just numbered 0-9, A-F.  however, on 3 of my M boards i have one entire bank at zero; 2 Bank 1's, and 1 Bank 4.  there must be some relationship btwn the 4 adjacent cards in those Banks.

any suggestions to revive those Banks?

Even v3 M-boards have 4 banks of 4 slots each.

Try reseating cards, or rotate the 4 cards within a bank.

If you think the problem is related to VRM thermal/current overload, try disabling a chip or two on each card, or turn down the voltage with the trimpots.

i have heatsinks and multiple fans so hopefully it isn't thermal.

i know the GUI shows 4 banks but do they have a function?  Dave did say they didn't mean anything.

Can you please post a screenshot of your bitfury webpage? Can you enable remote login to your rig for a few mins?

Since Bank1 and 2 are kaput, You will have to

a) Unplug all but first card from Bank 1 and Bank 2
b) Check if the cards on Bank 1 and  Bank 2 are mining
c) Add one more card to each Bank
d) Repeat b through c

what's the recommended procedure for doing this as in "stop miner" or "shut down completely" or "power off" before adding another card?

a) stop miner
b) shutdown -h now
c) power off PSU
d) PS no hot-swapping

[cypherdoc]

can someone explain the concept of the Banks and how they apply to v2.2 H boards and v3 M boards?

Dave said that the Banks are obsolete and that the H boards are just numbered 0-9, A-F.  however, on 3 of my M boards i have one entire bank at zero; 2 Bank 1's, and 1 Bank 4.  there must be some relationship btwn the 4 adjacent cards in those Banks.

any suggestions to revive those Banks?

Even v3 M-boards have 4 banks of 4 slots each.

Try reseating cards, or rotate the 4 cards within a bank.

If you think the problem is related to VRM thermal/current overload, try disabling a chip or two on each card, or turn down the voltage with the trimpots.

i have heatsinks and multiple fans so hopefully it isn't thermal.

i know the GUI shows 4 banks but do they have a function?  Dave did say they didn't mean anything.

Can you please post a screenshot of your bitfury webpage? Can you enable remote login to your rig for a few mins?

Since Bank1 and 2 are kaput, You will have to

a) Unplug all but first card from Bank 1 and Bank 2
b) Check if the cards on Bank 1 and  Bank 2 are mining
c) Add one more card to each Bank
d) Repeat b through c

what's the recommended procedure for doing this as in "stop miner" or "shut down completely" or "power off" before adding another card?

a) stop miner
b) shutdown -h now
c) power off PSU
d) PS no hot-swapping

and when a card shows 0 hashing, what's next?  decrease voltage with a quarter counterclockwise turn of the trimpot?

[Keefe]

what's the recommended procedure for doing this as in "stop miner" or "shut down completely" or "power off" before adding another card?

a) stop miner
b) shutdown -h now
c) power off PSU
d) PS no hot-swapping

I just do sudo poweroff, wait for all but the red light on the rPi to go off, and then shut off the PSU.

[Keefe]

and when a card shows 0 hashing, what's next?  decrease voltage with a quarter counterclockwise turn of the trimpot?

Have you already tried turning the speed down to 52 or 51? A few of my cards won't stay running at 53 or higher.

[cypherdoc]

what's the recommended procedure for doing this as in "stop miner" or "shut down completely" or "power off" before adding another card?

a) stop miner
b) shutdown -h now
c) power off PSU
d) PS no hot-swapping

I just do sudo poweroff, wait for all but the red light on the rPi to go off, and then shut off the PSU.

sounds reasonable to me as Chainminer seems broken.  i'd much rather work from the CLI.

[E]

how did you get bfgminer worked with your miner?  I was gave up bfgminer and I couldn't get it work with V1 mboard.

I install Raspbian and then follow the instruction here: https://bitcointalk.org/index.php?topic=251966.msg3530533#msg3530533

Then enable the the SPI with raspi-config before running bfgminer.

   Why hell blacklist spi

pi@raspberrypi ~ $ cat /etc/modprobe.d/raspi-blacklist.conf
# blacklist spi and i2c by default (many users don't need them)

#blacklist spi-bcm2708
blacklist i2c-bcm2708


I need to enable spi and I need to update raspi-config to see enable spi thing.

To get bfgminer working on my v1 m-board I made a quick change to

driver-bfsb.c


Change line 76 from:

Code:

        for (i = 0; i < 32; i++) {
                if (slot_on[i]) {
                        int chip_n;

to:

Code:

for (i = 0; i < 1; i++) {

As it was detecting all 96 chips in each slot and getting confused.

This isn't a proper patch, and you won't get per-card stats, but it does seem to work, and I can finally use eligius again w/o queue overruns

Code:

 bfgminer version 3.6.0 - Started: [2013-11-13 03:35:50] - [  0 days 00:07:13]
 [M]anage devices [P]ool management [S]ettings [D]isplay options  [H]elp [Q]uit
 Connected to stratum.mining.eligius.st diff 128 with stratum as user [redacted]
 Block: ...c42c35e1 #269332  Diff:511M ( 3.66Ph/s)  Started: [03:40:00]
 ST:18  F:0  NB:2  AS:0  BW:[247/419 B/s]  E:110.52  I: 6.34mBTC/hr  BS:214k
 1/96         | 199.1/187.3/179.7Gh/s | A:1243 R:0+0(none) HW:924/4.5%
--------------------------------------------------------------------------------
 BSB 0:       | 201.1/187.9/180.1Gh/s | A:1243 R:0+0(none) HW:926/4.5%
--------------------------------------------------------------------------------

bfgminer spends a lot of time reinitializing chips if their error rates are high; I'm not sure if this is slowing the whole thing down or not. Anyway, that's all the effort I'm putting into it for tonight

[cypherdoc]

and when a card shows 0 hashing, what's next?  decrease voltage with a quarter counterclockwise turn of the trimpot?

Have you already tried turning the speed down to 52 or 51? A few of my cards won't stay running at 53 or higher.

how is this done?

[Keefe]

what's the recommended procedure for doing this as in "stop miner" or "shut down completely" or "power off" before adding another card?

a) stop miner
b) shutdown -h now
c) power off PSU
d) PS no hot-swapping

I just do sudo poweroff, wait for all but the red light on the rPi to go off, and then shut off the PSU.

sounds reasonable to me as Chainminer seems broken.  i'd much rather work from the CLI.

Chainminer isn't just the web interface.

But anyway, here's the latest results from my manually-tuned v3 rig:
http://pastebin.com/raw.php?i=AM2rWAUg

I never adjusted the trimpots. I first turned off autotune and set all the speeds to 52, then rearranged cards a few times until they stopped cycling (or maybe the weather getting cooler did the trick). Maybe some slots get more airflow from my fans than others. Then I increased the speeds until cycling happened again and decreased the speeds for all chips on the cycling cards, increasing other cards' speeds until they cycled, etc. Once I got all cards stable, I began adjusting speeds of individual chips to find their sweet spots (speed vs error %). At one point, a card began cycling again, so I disabled one of the chips on that card that was producing half errors at any speed, to reduce the card's power usage and stop it from cycling.

[Keefe]

and when a card shows 0 hashing, what's next?  decrease voltage with a quarter counterclockwise turn of the trimpot?

Have you already tried turning the speed down to 52 or 51? A few of my cards won't stay running at 53 or higher.

how is this done?

Something like this:

Code:

rm /opt/bitfury/best.cnf ; for i in {1..256} ; do echo -e "$i\taIfDSo\t52" >> /opt/bitfury/best.cnf ; done

[cypherdoc]

can someone explain the concept of the Banks and how they apply to v2.2 H boards and v3 M boards?

Dave said that the Banks are obsolete and that the H boards are just numbered 0-9, A-F.  however, on 3 of my M boards i have one entire bank at zero; 2 Bank 1's, and 1 Bank 4.  there must be some relationship btwn the 4 adjacent cards in those Banks.

any suggestions to revive those Banks?

Even v3 M-boards have 4 banks of 4 slots each.

Try reseating cards, or rotate the 4 cards within a bank.

If you think the problem is related to VRM thermal/current overload, try disabling a chip or two on each card, or turn down the voltage with the trimpots.

i have heatsinks and multiple fans so hopefully it isn't thermal.

i know the GUI shows 4 banks but do they have a function?  Dave did say they didn't mean anything.

Not sure where Dave said that, but I have seen it mentioned by many people (dave including) that you want the fastest H card in the first slot of the bank, then order them in descending order with the slowest card in the last slot.

Can you link where he said it didn't matter, because I think it does and I dont think he'd say this. In practice, I can confirm its better to have the faster cards in the first slots of a bank otherwise it can cause the other cards to run slower or not run at all.

"Don't pay attention to the Bank 1, Bank 2, etc.  Those line up for V2 M-boards.  The numbers of the slots do line up as 1-16 map to 1 - 9, A - F.  There, you can see the output of each H-card."

https://bitcointalk.org/index.php?topic=288109.0;all

[CoinHoarder]

Hmm, wish you could quote it in context.. anyways.. Anyone else having trouble getting bfgminer built? I've tried it how I thought it should be done, and a couple instructions of how others thought it should be done. Nothing's working.

This is so frustrating.. after over 24 hours of troubleshooting, one is now running at full speed on admin page, but half speed on the pool. The other I seem to have corrupted chainminer while troubleshooting it and it isn't hashing at all. Certainly buzzdave is experiencing these same issues since he is using the same hardware to setup the 200 Th mine. I wish he would be a little bit more proactive about supporting us... perhaps release an updated firmware that fixes a lot of the issues everyone's having. Usually when I spend $16,000 on something, I expect it to work.

[Doff]

I'm confused as to what the A autotune is supposed to do, does it not raise it past 53? Im getting similar numbers to Keefe but have everything just running with Autotune at the moment.

[Doff]

Hmm, wish you could quote it in context.. anyways.. Anyone else having trouble getting bfgminer built? I've tried it how I thought it should be done, and a couple instructions of how others thought it should be done. Nothing's working.

This is so frustrating.. one is now running at full speed on admin page, but half speed on the pool. The other I seem to have corrupted chainminer while troubleshooting it and it isn't hashing at all. Certainly buzzdave is experiencing these same issues since he is using the same hardware to setup the 200 Th mine. I wish he would be a little bit more proactive about supporting us... perhaps release an updated firmware that fixes a lot of the issues everyone's having. Usually when I spend $16,000 on something, I expect it to work.

Did you adjust your Difficulty for the pool for one running at half on the pool?

[CoinHoarder]

Did you adjust your Difficulty for the pool for one running at half on the pool?

The only pools I've tried... Eligius and Slush don't have adjustable difficulty I don't think.

Maybe I should try one where I can..

[Xian01]

Hope to do some troubleshooting of the second rig tonight after picking up some proper tools at Frys earlier.

Is it safe to assume that turning the pot all the way left should be most stable, and turn all the way to the right at your peril ?

 Also, thanks Frys dude for selling me a Netgear WNDR3400v3 and stupid me trusting him that it works with Dd-wrt for use as a wireless bridge...

[Xian01]

Did you adjust your Difficulty for the pool for one running at half on the pool?

The only pools I've tried... Eligius and Slush don't have adjustable difficulty I don't think.
Maybe I should try one where I can..

Ghash.io is super spiffy that way, and btcguild is really good at automatically setting your diff to handle your hash rate, in my experience.

[Doff]

Did you adjust your Difficulty for the pool for one running at half on the pool?

The only pools I've tried... Eligius and Slush don't have adjustable difficulty I don't think.

Maybe I should try one where I can..

Yeah, these are pretty sensitive to that.

[CoinHoarder]

Did you adjust your Difficulty for the pool for one running at half on the pool?

The only pools I've tried... Eligius and Slush don't have adjustable difficulty I don't think.

Maybe I should try one where I can..

Yeah, these are pretty sensitive to that.

Wow.. I switched over to BTCGuild @ 512 difficulty & it is already showing faster than both Eligius and Slush's pool. It's showing 476, hopefulyl it winds up closer to the 524 Gh the firmware shows.

I'm going to cross my fingers that this was what was wrong all along. I'm so happy I made a backup image of the sdcard before I started troubleshooting, I can restore that image to fix the corrupted chainminer and hopefully be good to go if the pool was really the issue.

protip: backup your sd card using a digital camera/laptop with sdcard reader/sdcard reader and use win32diskmanager to copy the image file from the sdcard to your computer for a back up. You can use the same tool to write the image to the sdcard as well. The images provided in the bitfury support guide thread (in mining support subforum) were too large to fit on my sdcard for some reason.

[greaterninja]

i recommend signup for a free account at ghash.io