Hacking The KNC Firmware: Overclocking

[joeventura]

Ok I helped all you guys out, see if you can assist.

I just added a 4th module to make a Jupiter and of course with my luck it has a Dead Die 0

I am overclocked to 231

As soon as I try 241 or 251 dies start getting shut down due to HW errors.

As you can see I have the temperature covered.

Whats your suggestion on settings? What should I adjust in advanced?

This is a October Saturn/Jupiter with 4 VRMs

Thanks for any suggestions

Have you not tried increasing the voltage just on that die? Works for me

Yes it does not respond sadly

[1713576977]

1713576977

[1713576977]

1713576977

[elasticband]

Ok I helped all you guys out, see if you can assist.

I just added a 4th module to make a Jupiter and of course with my luck it has a Dead Die 0

I am overclocked to 231

As soon as I try 241 or 251 dies start getting shut down due to HW errors.

As you can see I have the temperature covered.

Whats your suggestion on settings? What should I adjust in advanced?

This is a October Saturn/Jupiter with 4 VRMs

Thanks for any suggestions

Have you not tried increasing the voltage just on that die? Works for me

Yes it does not respond sadly

Have you tried to INCREASE the temperatures?

edit: this is how i would go about reviving a dead DIE, just watch, once it kicks in temps can increase quite fast in my experience and cooling them to the levels you are at just now will just kill it again.

all my DIE 1 which have been dead are revived by changing to 3.3 spi volt, but this will kill the others until the reach a certain temperature

[DanBUK]

Thank you, to the various posters within this thread. I think i've managed to squeeze:

Oct Saturn: from 280GH/s -> 320GH/s
Nov Jupiter: from 660GH/s -> 700GH/s

Admittedly only for a few hours. But temps, draw etc are looking quite stable.

I've seen a number of people bash others for asking simple questions. Reading the entirety of this thread takes time! A lot of time (be honest, there is a lot within). I just want to throw it out there that bashing people for not reading a +30page thread (30x20 = 600 replies) is a little eager.

Suggestion: we move the OC of KnC devices onto a wiki style page, thereby the information can be easily updated, and formed into a readable form. (Unlike ~600 fragments that we all need to interpret and join together each time a new person reads?)

//rant// (Don't even get me started on the failings of people to understand what vi is, and how to use it. They don't deserve the wiki page... (If you don't understand bash, then you shouldn't be logged into a *nix machine... Let alone be allowed to use vi. Or attempt to OC hardware that you don't fully understand)) //end rant//

Edit: (Maybe I was a little OTT with my rant, but if you read some advice from someone, they tell you to use vi (or any tool), and then when you don't get it you don't do any research yourself, just ask again in the forum/post, then that's lazy.)

[mekadeka]

Suggestion: we move the OC of KnC devices onto a wiki style page, thereby the information can be easily updated, and formed into a readable form. (Unlike ~600 fragments that we all need to interpret and join together each time a new person reads?)

I second this, and I will personally donate a little to someone who gets this done. Some criteria:

It will need to be a top-down guide for lay users.
It will need to include step-by-step instructions.
It will need to include correct order of procedure so that updates and/or configuration won't wipe out anything.
It should include persistent bertmod, enablecores (if needed), wizkid's cgminer,etc.

I would be willing to edit it for clarity, at least for some reasonable amount of time going forward.

Any takers?

[temen]

Hi

How it's possible to make script that copies my own cgminer.sh script to /etc/init.d/cgminer.sh on startup?

And how to make script that restarts the miner with like 3 hour intervals? As there has been some downtime with overclocking and unit just shutting down=)

now 500gh/s semi-stable 2 module november saturn, was on during the night. Started clocking each die individually. To make it stable i needed to get rid of those fault 4:s turning up after some time. I think they have timer on them and that causes some overtemp-protection to trigger in after some time.

Values for asic 4 are 361 and asic 0 varies in between 361 and 2C1.


Description=cgminer 3.9.0|SUMMARY

MHS av=503541.30
MHS 5s=503582.25

Work Utility=7034.48
Difficulty Accepted=245831.00000000
Difficulty Rejected=1248.00000000
Difficulty Stale=0.00000000
Best Share=376240
Device Hardware%=0.1790
Device Rejected%=0.5065
Pool Rejected%=0.5051
Pool Stale%=0.0000|

ASIC Board    Info
0    

Temperature sensor: 46.0 C
Die ID    Cores ON    Cores OFF    %
0   48   0    100
1   48   0    100
2   48   0    100
3   48   0    100
DC/DC ID    ON/OFF    Status    Input Voltage    Output Voltage    Output Current
0    OFF    OK    11.5 V    0.82 V    35.2 A (28.9 W)
1    OFF    OK    11.4 V    0.822 V    34.6 A (28.4 W)
2      OFF    OK    11.3 V    0.815 V    37.5 A (30.6 W)
3    OFF    OK    11.4 V    0.817 V    37.7 A (30.8 W)
4    OFF    OK    11.3 V    0.821 V    35.9 A (29.5 W)
5    OFF    OK    11.4 V    0.818 V    36.1 A (29.5 W)
6    OFF    OK    11.3 V    0.826 V    29.5 A (24.4 W)
7     OFF    OK    11.3 V    0.82 V    30.2 A (24.8 W)
4    

Temperature sensor: 47.5 C
Die ID    Cores ON    Cores OFF    %
0   48   0    100
1   48   0    100
2   48   0    100
3   48   0    100
DC/DC ID    ON/OFF    Status    Input Voltage    Output Voltage    Output Current
0    OFF    OK    11.5 V    0.82 V    38.2 A (31.3 W)
1    OFF    OK    11.4 V    0.821 V    37.8 A (31 W)
2    OFF    OK    11.5 V    0.822 V    37.4 A (30.7 W)
3    OFF    OK    11.5 V    0.817 V    37.6 A (30.7 W)
4    OFF    OK    11.5 V    0.821 V    37.5 A (30.8 W)
5    OFF    OK    11.4 V    0.816 V    37.8 A (30.8 W)
6    OFF    OK    11.5 V    0.816 V    36.6 A (29.9 W)
7    OFF    OK    11.4 V    0.817 V    36.6 A (29.9 W)

Total DC/DC power output: 471 W

vrm's for asic 0 die 3 make fault 4 risen above that 30.  Asic 0 is heating more than 4 in overall.

[ncs0ne]

Suggestion: we move the OC of KnC devices onto a wiki style page, thereby the information can be easily updated, and formed into a readable form. (Unlike ~600 fragments that we all need to interpret and join together each time a new person reads?)

I second this, and I will personally donate a little to someone who gets this done. Some criteria:

It will need to be a top-down guide for lay users.
It will need to include step-by-step instructions.
It will need to include correct order of procedure so that updates and/or configuration won't wipe out anything.
It should include persistent bertmod, enablecores (if needed), wizkid's cgminer,etc.

I would be willing to edit it for clarity, at least for some reasonable amount of time going forward.

Any takers?

Regarding this wiki-style page...how to realize ?

I'll contribute, but I'd prefer to have all donations flowing to those who risked their hw and shared their knowledge with us in order to motivate such people for the next time we might need them. Writing a guide is no work/value in comparison.

I think it was tolip_wen and temen, in case I forgot one please forgive me and stand up.

Hi

How it's possible to make script that copies my own cgminer.sh script to /etc/init.d/cgminer.sh on startup?

And how to make script that restarts the miner with like 3 hour intervals? As there has been some downtime with overclocking and unit just shutting down=)

now 500gh/s semi-stable 2 module november saturn, was on during the night. Started clocking each die individually. To make it stable i needed to get rid of those fault 4:s turning up after some time. I think they have timer on them and that causes some overtemp-protection to trigger in after some time.

To have a custom script executed at boot time, you need to modify/rebuild the firmware (I'm going to do so for myself these days as soon as I'm done I'll post it in here how to do it)
Regarding your 3h question, a simple cron job or some kind of loop with sleeps would do the trick in general.

[DanBUK]

Regarding this wiki-style page...how to realize ?

I'm going to make a start on a page on http://en.bitcoin.it/ today. Will let you all know once I've made some progress.

[sickpig]

I must have missed the code on how to change OC on a die by die basis.

How can that be done?

Did you happen to find it?

I'm in a hurry now but If memory serves it's only a matter to change zzz.sh in such a way to issue different command dies:

Code:

cmd=$(printf "0x86,0x%02X,0x02,0x11" $c)
spi-test -s 50000 -OHC -D /dev/spidev1.0 $cmd >/dev/null

$c is a convenient variable that could takes value from 0 to 3, one for each die of the slot.

As soon as I've time I'll share a modified version that let you tweak the OC die by die.  

[joeventura]

I must have missed the code on how to change OC on a die by die basis.

How can that be done?

Did you happen to find it?

I'm in a hurry now but If memory serves it's only a matter to change zzz.sh in such a way to issue different command dies:

Code:

cmd=$(printf "0x86,0x%02X,0x02,0x11" $c)
spi-test -s 50000 -OHC -D /dev/spidev1.0 $cmd >/dev/null

$c is a convenient variable that could takes value from 0 to 3, one for each die of the slot.

As soon as I've time I'll share a modified version that let you tweak the OC die by die.  

Thanks I look forward to that!

[CYPER]

https://bitcointalk.org/index.php?topic=313978.msg4432410#msg4432410

[mekadeka]

Regarding this wiki-style page...how to realize ?

I'll contribute, but I'd prefer to have all donations flowing to those who risked their hw and shared their knowledge with us in order to motivate such people for the next time we might need them. Writing a guide is no work/value in comparison.

I think it was tolip_wen and temen, in case I forgot one please forgive me and stand up.

To have a custom script executed at boot time, you need to modify/rebuild the firmware (I'm going to do so for myself these days as soon as I'm done I'll post it in here how to do it)
Regarding your 3h question, a simple cron job or some kind of loop with sleeps would do the trick in general.

Well, I and others have been running their machines with the same risks as everyone else and have tested and contributed to the best of their knowledge/capabilities.

Also, I don't think the value of a comprehensive, exact guide should be devalued at all. Compiling knowledge, findings, and techniques into a centralized, valuable, and exact resource is just as useful as any other contribution, if not more so. This isn't about the act of writing, it's about someone acting as an expert editor of a knowledgebase. That's pretty significant stuff.

[rmarleal]

October Jupiter 4 vrm. 251

716 Ghs.

[CeeCee]

October Jupiter 4 vrm. 251

716 Ghs.

[jelin1984]

Do you have extra fan at October Jupiter??

How many days you hashing at that speed?

[idee2013]

Do you have extra fan at October Jupiter??

How many days you hashing at that speed?

my around 2 weeks

[jelin1984]

At 251?

Any extra cooler fan? What was the increase at watts?

[idee2013]

At 251?

Any extra cooler fan? What was the increase at watts?

sorry but are you stupid?

how is the extra watt after overclocked the jupiter?

does anyone know?

do you read any previous posts? you ask very often things that were already answered.

again??? ...read the posts!!!!!!!!!!!

i read but only see overclocked for octber jupiter not for november jupiter

so you dont have read the thread!

reaching now 12h avg. 715 at pool with 1x241 (8vrm) and 3x251(2x 8vrm+ 4x vrm) oct.

reaching now 12h avg. 715 at pool with 1x241 (8vrm) and 3x251(2x 8vrm+ 4x vrm) oct.

nice!

what amps/temps do you get?

very mixed volt 0,09 - 0,13, amps 55-59, spi 3,1v spi clock 307200 Hz,
temps 38°C-55°C, with heatsinks on vrms and with a tower big fan blowing from the top to the bottom of the case...the airflow is splited than under and over the asic boards...
hw 0,8-1,1%  r 0,05% WU ~9870

use one like that http://aquanaturalzone.com/images/t2.jpg

it is laying  at the top and the middle of the miner with medium blowingpower level

so every part is "only" warm

847 watt on the wall (without big fan) with P&C 1200 MK III platinum
99.2 FW using cgminer

enough information?  

hi, for my jupiter 4vrm October with OC to 201 with 211 going well and gives less performance that has 201? when would have to give more power.
the firm is 0.99.1tune.

regards

help, because 231 have much less power than a 201? as I can get to 670-680GH?

October 4 asics jup

help please

hi,

https://bitcointalk.org/index.php?topic=313978.msg4425943#msg4425943

have 99.1 tune, too

4vrm asic dies -0,1003 volts each
each VRM got  VRAM heatsinks and i have a big fan over the Jupiter

...and much more in this thread...

[sickpig]

https://bitcointalk.org/index.php?topic=313978.msg4432410#msg4432410

@joeventura this precisely what you need to do.

thanks CYPER

[rmarleal]

I have been running it at 251 for a couple of weeks with no problems. Temperatures are under control with open case, frontal fans in 45º on the frontal asics, y a big fan in the top of the case, heatsinks and artic cooler grease on the vrms.
3.3 volts at spi and around 60 amp in all dies.

716 ghs with 10006 WU. At the pool 740ghs.

[DPoS]

I have been running it at 251 for a couple of weeks with no problems. Temperatures are under control with open case, frontal fans in 45º on the frontal asics, y a big fan in the top of the case, heatsinks and artic cooler grease on the vrms.
3.3 volts at spi and around 60 amp in all dies.

716 ghs with 10006 WU. At the pool 740ghs.

was that 4VRM Oct boards?