Pages:
Author

Topic: [GUIDE] BitFury Miner Support/Tuning - page 38. (Read 148018 times)

legendary
Activity: 1946
Merit: 1006
Bitcoin / Crypto mining Hardware.
September 15, 2013, 01:39:55 AM
#93


Here's How I did it:

a) Buy Copper VGA RAM Heatsinks. Usually ships in sets of 8, so 2 packs per h-board.

a.1) I bought these:  http://www.ebay.com/itm/161102078777?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649

b) Affix each piece of VGA RAM Heatsink to the back of each BitFury chip on the H-Board. The RAM heatsinks come with thermal adhesives affixed.

c) Cut a 2 inch strip of DDR3 / DDR2 RAM heatskink for cooling the regulator area. Affix the heatsink to the back of board near the regulator with either thermally conductive glue or use some kind of thermal paste and small dabs of superglue / epoxy at the edges

d) Pencil mod the board until voltage reaches 0.8V. One of my boards is at 0.8V another at 0.84V. I would say 0.85V is the max after which either the regulator will shut down or the hash rate will drop dramatically.

e) Cool the boards with 120mm or bigger cooling fans.

f) Play around with best.cnf (I first changed of all the chip speeds to 54, and measured nonce rate after 10mins of hashing, then changed to 55 and again measured nonce rates. Compared the differences on a excel spreadsheet, and chose 54 or 55 depending on which produced higher hash rates for the chip.)

g) I Use only one pool for mining

h) renice the miner proxy process to -14 or lower

Code:
2100 root        6 -14 33196 18864  4064 R 81.0  3.8 36h27:30 python ./mining_proxy.py -o us3.eclipsemc.com -p 3333 -cu XXXXXXXXXXX -cp XXXX -gp 8332 -v
h.1) I use htop to do it
h.1.1) sudo apt-get install htop
h.1.2) pi@bitfury / $ sudo htop
h.1.3) Press F7 repeatedly to reduce nice.
        


hero member
Activity: 525
Merit: 500
..yeah
September 14, 2013, 04:08:01 PM
#92

now how did you manage that? Crazy. Due to better cooling or mostly due to massive overclocking, pencil mod it down to .. 1k? 1,05k? Of course fine tuning in the end.. but.. 38? damn!
cet
member
Activity: 100
Merit: 10
September 14, 2013, 01:14:42 PM
#91
What's the solution when you see lots of MISO errors?

by lots I mean:
Code:
Last login: Sat Sep 14 17:51:00 2013 from 192.168.1.2
{ "stats":
 {"speed": 880, "noncerate": 21.589, "noncerateperchip":1.349, "hashrate":28.678
, "good":568, "errors":138, "spi-errors":0, "miso-errors":126, "jobs":140, "record":0.000
,"boards": [
{ "slot": "0", "speed": 880, "noncerate":21.589, "hashrate": 28.678, "good": 568
, "errors": 138, "spi-errors": 0, "miso-errors":[b]126[/b] }
]
  } }
/cet
member
Activity: 94
Merit: 10
September 14, 2013, 09:50:54 AM
#90
That is chip 2, and the jumper should not be closed.  This could be the problem with your board. The chips are numbered U40, U41, ..., so chip 9 would be U48.

thank you!

*edit*
removing the solder from SJ50 helped alot, miso-errors on other chips are gone and hashrate doubled  Smiley
hero member
Activity: 574
Merit: 501
September 14, 2013, 09:49:08 AM
#89
That is chip 2, and that jumper should not be closed.  This could be the problem with your board. The chips are numbered U40, U41, ..., so chip 9 would be U48.
member
Activity: 94
Merit: 10
September 14, 2013, 09:41:23 AM
#88
plz ... feel free to correct me or give some suggestions ...

I dont think you are talking about a real bad chip here, look at my chip 9 for example it only does errors ever, no point in turning it on at all.

Code:
1       aIfDSo  52      1.346   1.289   94      0       0       0       122     [8:0]   0       6 6 6 6 6 6 6 6 6 6 5 5 6 6 6 6         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
2       aIfDSo  52      1.160   1.311   81      2       0       0       124     [8:1]   36      5 6 6 5 5 4 5 5 5 5 5 5 5 5 5 5         0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0
3       aIfDSo  52      1.274   1.321   89      3       0       0       125     [8:2]   0       6 6 5 5 6 6 6 6 5 6 6 6 5 5 5 5         0 0 1 1 0 0 0 0 1 0 0 0 0 0 0 0
4       aIfDSo  52      1.374   1.216   96      0       0       0       115     [8:3]   0       6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
5       aIfDSo  52      1.160   1.004   81      0       0       0       95      [8:4]   15      5 5 5 5 5 6 5 5 5 5 5 5 5 5 5 5         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
6       aIfDSo  52      1.174   1.374   82      0       0       2       130     [8:5]   36      5 6 6 5 5 5 5 5 5 5 5 5 5 5 5 5         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
7       aIfDSo  52      1.346   1.332   94      1       1       0       126     [8:6]   0       6 6 5 6 6 6 6 6 6 6 6 6 6 6 5 6         0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0
8       aIfDSo  52      1.503   1.300   105     3       0       0       123     [8:7]   1       6 5 6 7 7 7 7 7 7 7 7 7 6 6 7 6         0 1 0 0 0 0 0 0 0 0 0 0 1 1 0 0
9       aIfDSo  0       0.000   0.285   0       2208    0       6       27      [8:8]   756     0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0         0 315 316 316 0 316 315 0 0 0 0 0 0 0 315 315
10      aIfDSo  52      1.389   1.226   97      4       0       4       116     [8:9]   10      6 6 6 6 5 5 5 6 5 7 7 7 6 7 7 6         0 0 0 0 1 2 1 0 0 0 0 0 0 0 0 0
11      aIfDSo  0       0.000   0.888   0       85      0       7       84      [8:A]   495     0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0         5 5 5 5 5 6 6 6 6 6 5 5 5 5 5 5
12      aIfDSo  52      0.000   1.247   0       92      0       4       118     [8:B]   23      0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0         6 6 6 6 5 6 7 6 5 4 6 6 6 6 5 6
13      aIfDSo  52      0.000   1.237   0       75      0       1       117     [8:C]   147     0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0         4 4 5 5 5 5 5 5 5 5 4 4 5 5 4 5
14      aIfDSo  52      0.000   1.226   0       0       0       0       116     [8:D]   13      0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
15      aIfDSo  52      1.088   1.332   76      5       0       2       126     [8:E]   0       4 5 4 5 6 3 5 5 6 6 6 5 5 4 4 3         1 0 1 0 0 1 1 0 0 0 0 0 0 1 0 0
16      aIfDSo  52      0.129   0.285   9       1       0       7       27      [8:F]   53      0 0 1 1 0 0 1 1 1 1 1 1 1 0 0 0         0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0
17      aIfDSo  52      0.000   0.190   0       0       0       0       18      [C:0]   194     0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
18      aIfDSo  52      1.303   1.374   91      1       0       0       130     [C:1]   0       4 6 6 6 6 6 6 6 6 6 6 6 6 5 5 5         1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
19      aIfDSo  52      1.016   1.279   71      1       0       0       121     [C:2]   0       4 4 5 5 5 4 5 5 5 5 4 4 4 4 4 4         0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0
20      aIfDSo  52      1.589   1.342   111     0       0       0       127     [C:3]   0       7 7 7 7 6 7 7 7 7 7 7 7 7 7 7 7         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
21      aIfDSo  52      0.873   1.321   61      0       0       0       125     [C:4]   0       4 4 4 4 4 4 4 4 4 4 4 4 4 3 3 3         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
22      aIfDSo  52      0.000   1.205   0       88      0       24      114     [C:5]   15      0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0         6 5 6 6 5 6 6 5 5 6 6 5 5 5 5 6
23      aIfDSo  0       0.000   0.000   0       0       0       0       0       [C:6]   573     0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
24      aIfDSo  52      0.000   0.000   0       0       0       0       0       [C:7]   294     0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
25      aIfDSo  52      0.000   0.000   0       0       0       0       0       [C:8]   216     0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
26      aIfDSo  52      0.000   0.000   0       0       0       0       0       [C:9]   49      0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
27      aIfDSo  0       0.029   0.095   2       8       0       0       9       [C:A]   665     0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0         0 0 0 0 0 1 1 1 1 1 1 0 1 0 1 0
28      aIfDSo  52      1.374   1.279   96      2       0       0       121     [C:B]   10      5 7 6 5 4 6 6 7 7 7 6 6 5 7 6 6         0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0
29      aIfDSo  52      0.601   1.247   42      0       0       0       118     [C:C]   4       3 2 2 2 3 2 3 3 3 3 3 3 3 2 2 3         0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
30      aIfDSo  52      1.260   1.226   88      6       0       0       116     [C:D]   2       5 5 5 5 5 5 6 6 6 6 5 6 6 6 6 5         1 2 0 0 1 0 0 0 0 0 1 0 0 0 0 1
31      aIfDSo  52      0.959   1.184   67      2       0       1       112     [C:E]   1       5 4 4 5 4 4 4 4 3 4 4 4 5 4 4 5         0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0
32      aIfDSo  52      1.102   1.300   77      4       0       0       123     [C:F]   0       4 5 5 6 5 4 4 4 6 6 6 6 4 4 3 5         0 1 1 0 0 1 0 1 0 0 0 0 0 0 0 0
speed:1456 noncerate[GH/s]:23.050 (0.720/chip) hashrate[GH/s]:30.916 good:1610 errors:2591 spi-errors:1 miso-errors:58 jobs:376 (record[GH/s]:28.862)
8:      728     12.942  17.873  904     2479    1       33
C:      728     10.107  13.043  706     112     0       25

after chip #9 every coming chips go bad too, so I think I have to use the soldering jumpers - but I dont know how. Also it seems someone already used the soldering jumper on my chip #9 (if it is nr. 9). Can someone advice me here?

sr. member
Activity: 434
Merit: 265
September 14, 2013, 05:31:09 AM
#87
added ...

6. Bad & Dead Chips

You might have dead chips (marked as A) on your board that want go anywhere else then 0, ... and autotune ("a") will even be turn of on them and you might have bad chips which autotune will probably tune down to 0 clockspeed, but still autotune ("A") will be turned on (marked as B).



For dead chips there is no solution from software side that I know off. There might be some Hardware possibilities ... I don't know.

For bad chips there is!

First you need to turn of autotune "A" to "a" so that they wont be turned down to 0, because right know as long as they produce to many errors, they will be turned down until 0.

To start of set there clockspeed to 55 on the bad chip and let him run for 10-15 min. Check again. If he is performing better or not ... find a good rate between good and error. You might go down to 54 or 53 ... just check again, where the bad chip is performing at his max. potential ... at the end it might look like this one ...


* a bad one but still more then a stick

It might even be that bad chips get again running fine after running for hours (>24h) ... some users mentioned that bitfury chips have an integrated selfhealing mechanisme ... :-)

plz ... feel free to correct me or give some suggestions ...
sr. member
Activity: 434
Merit: 265
September 14, 2013, 04:21:57 AM
#86
yep ... you might post ... your modifications ... ^^
GH
member
Activity: 117
Merit: 10
September 14, 2013, 03:36:47 AM
#85
Can you guess the noncerate of the card within +-1GH/s ?


38 GH .. ?


Wink
right on


Nice! But still inside the specs of the DC/DC Converter? Hard to belive for me, assumed that cooling alone does not produce more nonces  Grin
legendary
Activity: 1946
Merit: 1006
Bitcoin / Crypto mining Hardware.
September 14, 2013, 02:43:29 AM
#84
Can you guess the noncerate of the card within +-1GH/s ?


38 GH .. ?


Wink
right on
sr. member
Activity: 434
Merit: 265
September 14, 2013, 02:27:00 AM
#83
Can you guess the noncerate of the card within +-1GH/s ?


38 GH .. ?
legendary
Activity: 1946
Merit: 1006
Bitcoin / Crypto mining Hardware.
September 14, 2013, 01:17:42 AM
#82
Can you guess the noncerate of the card within +-1GH/s ?
legendary
Activity: 1946
Merit: 1006
Bitcoin / Crypto mining Hardware.
September 14, 2013, 01:14:18 AM
#81
sr. member
Activity: 408
Merit: 250
September 14, 2013, 01:02:29 AM
#80
I'm dying to know when Bitfury will be more open and finally let others do what they do best: mining software without imposing them support restrictions and conditions. Not sure if there's any politics involved or what's the contingency for not doing so.
What's not open about bitfury ? All the information needed to port other mining software to the Raspberry PI is available.

Well, there's perhaps a confusion here but I'm almost sure I read cgminer's developing team sort of complaining about some lack of collaboration or difficulty to let cgminer's mining software to be available for bitfury and that they were asked to pay for support or the like.
I just really look forward to seeing more available and tested mining software just like with most other mining hardware rather than sort of proprietary software. Feel free to correct me but for some reason that doesn't seem to be the case with Bitfury.
sr. member
Activity: 251
Merit: 250
September 14, 2013, 12:32:06 AM
#79
I'm dying to know when Bitfury will be more open and finally let others do what they do best: mining software without imposing them support restrictions and conditions. Not sure if there's any politics involved or what's the contingency for not doing so.
What's not open about bitfury ? All the information needed to port other mining software to the Raspberry PI is available.
sr. member
Activity: 408
Merit: 250
September 13, 2013, 09:01:04 PM
#78
cgminer builds ok
most of the display options are broken though and hang the miner

Code:

BITFURY 0: 48.98G/49.77Gh/s | A:1408 R:0 HW:0 WU:697.5/m
 
[2013-09-13 10:41:37] Accepted 07c8aaf3 Diff 32/32 BITFURY 0
 [2013-09-13 10:41:41] Accepted 05d8fca3 Diff 43/32 BITFURY 0
 [2013-09-13 10:41:47] Accepted 001f09a1 Diff 2.11K/32 BITFURY 0
 [2013-09-13 10:41:49] Accepted 0004a45a Diff 14.1K/32 BITFURY 0
 [2013-09-13 10:41:51] Accepted 025c7689 Diff 108/32 BITFURY 0
 [2013-09-13 10:41:52] Accepted 06cd424b Diff 37/32 BITFURY 0
 [2013-09-13 10:41:54] Accepted 07ae8f6b Diff 33/32 BITFURY 0
 [2013-09-13 10:41:54] Accepted 043e8079 Diff 60/32 BITFURY 0

tbh its nice, but think ill wait for bfgminer with auto adjusting freq, Im not keen on cgminer myself.

I'm dying to know when Bitfury will be more open and finally let others do what they do best: mining software without imposing them support restrictions and conditions. Not sure if there's any politics involved or what's the contingency for not doing so.

Cgminer or Bfgminer please come to Bitfury...or viceversa!!  :p
 
sr. member
Activity: 658
Merit: 250
September 13, 2013, 06:45:06 PM
#77
I had even higher rates with a faulty chip which recovered for 24 hours and didn't throw errors, what it now does again after a restart...?

I have a similar chip. It takes around 24 hours until it stops producing errors after a poweroff & restart.
GH
member
Activity: 117
Merit: 10
September 13, 2013, 06:09:37 PM
#76
Just to let you know what is possible:

speed:1760 noncerate[GH/s]:63.651 (1.989/chip) hashrate[GH/s]:67.191 good:4446 errors:207 spi-errors:1 miso-errors:0 jobs:375 (record[GH/s]:65.097)
4:   880   32.985   33.643   2304   42   0   0
8:   880   30.666   33.548   2142   165   1   0
                   
This is NOT the best result, just a momentary (average) snapshot. Overvolting and manual tuning (resulting speeds all 55 here) will do wonders with this great (but sometimes strange, especially when cold) hardware!

I had even higher rates with a faulty chip which recovered for 24 hours and didn't throw errors, what it now does again after a restart...?

Good luck and thanks for the input and efforts of everybody here!

 Grin

sr. member
Activity: 350
Merit: 250
September 13, 2013, 06:43:04 AM
#75
cgminer builds ok
most of the display options are broken though and hang the miner

Code:

BITFURY 0: 48.98G/49.77Gh/s | A:1408 R:0 HW:0 WU:697.5/m


[2013-09-13 10:41:37] Accepted 07c8aaf3 Diff 32/32 BITFURY 0
 [2013-09-13 10:41:41] Accepted 05d8fca3 Diff 43/32 BITFURY 0
 [2013-09-13 10:41:47] Accepted 001f09a1 Diff 2.11K/32 BITFURY 0
 [2013-09-13 10:41:49] Accepted 0004a45a Diff 14.1K/32 BITFURY 0
 [2013-09-13 10:41:51] Accepted 025c7689 Diff 108/32 BITFURY 0
 [2013-09-13 10:41:52] Accepted 06cd424b Diff 37/32 BITFURY 0
 [2013-09-13 10:41:54] Accepted 07ae8f6b Diff 33/32 BITFURY 0
 [2013-09-13 10:41:54] Accepted 043e8079 Diff 60/32 BITFURY 0

tbh its nice, but think ill wait for bfgminer with auto adjusting freq, Im not keen on cgminer myself.
sr. member
Activity: 408
Merit: 250
September 13, 2013, 12:50:30 AM
#74
Most of the stratum log lines I read are asking for new work and barely are there with "accepted work" . It used to list a lot of "accepted" share lines but not anymore. Why the change? Is that a problem???

Code:
NFO proxy getwork_listener._on_authorized # Worker 'XXXXX' asks for new work
2013-09-13 04:41:53,612 INFO proxy jobs.submit # Submitting 432aa92
2013-09-13 04:41:53,617 DEBUG protocol protocol.writeJsonRequest # < {"params": ["XXXXX", "1372565647 11709", "00621839", "585297df", "f198c1a7"], "id": 3838, "method": "mining.submit"}
2013-09-13 04:41:53,675 INFO proxy getwork_listener._on_authorized # Worker 'XXXXX' asks for new work
2013-09-13 04:41:53,693 DEBUG protocol protocol.lineReceived # > {u'error': None, u'result': True, u'id': 3591}
2013-09-13 04:41:53,697 WARNING proxy getwork_listener._on_submit # [75ms] Share from 'XXXXX' accepted, diff 247
2013-09-13 04:41:53,803 INFO proxy getwork_listener._on_authorized # Worker 'XXXXX' asks for new work
2013-09-13 04:41:53,930 INFO proxy getwork_listener._on_authorized # Worker 'XXXXX' asks for new work
2013-09-13 04:41:54,057 INFO proxy getwork_listener._on_authorized # Worker 'XXXXX' asks for new work
2013-09-13 04:41:54,183 INFO proxy getwork_listener._on_authorized # Worker 'XXXXX' asks for new work
2013-09-13 04:41:54,311 INFO proxy getwork_listener._on_authorized # Worker 'XXXXX' asks for new work
2013-09-13 04:41:54,437 INFO proxy getwork_listener._on_authorized # Worker 'XXXXX' asks for new work

My pool seems to be reporting around my correct hashrate. Also, I checked my .putstat.log and it shows this:

Code:
pi@bitfury ~ $ cat /run/shm/.putstat.log
0 65693 3169 3169 [0]http://127.0.0.1:8332/
0 0 0 0 [1]http://127.0.0.1:8333/
0 0 0 0 [2]http://127.0.0.1:8334/


Do you see any problems??  Please help  Sad

Pages:
Jump to: