KnCMiner Jupiter Miner First Impressions

[jelin1984]

MINE WITH 0.91 FIRMWARE

ASIC Board    Info
0    

Temperature sensor: 44.5 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    ON    OK    11.9 V    0.71 V    39.4 A (28 W)
1    OFF    OFF    12 V    0.706 V    0 A
2    ON    OK    11.8 V    0.753 V    40.1 A (30.2 W)
3    OFF    OFF    11.9 V    0.736 V    0 A
4    ON    OK    12 V    0.735 V    38.1 A (28 W)
5    OFF    OFF    11.9 V    0.703 V    0 A
6    OFF    OFF    12 V    0.741 V    0 A
7    ON    OK    12 V    0.755 V    40.6 A (30.7 W)
1    

Temperature sensor: 56.0 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    ON    OK    11.9 V    0.735 V    39.7 A (29.2 W)
1    OFF    OFF    11.9 V    0.712 V    0 A
2    ON    OK    12 V    0.711 V    41.4 A (29.4 W)
3    OFF    OFF    11.9 V    0.741 V    0 A
4    ON    OK    12 V    0.722 V    40.8 A (29.5 W)
5    OFF    OFF    12 V    0.71 V    0 A
6    OFF    OFF    12 V    0.729 V    0 A
7    ON    OK    12 V    0.735 V    39.7 A (29.2 W)
2    

Temperature sensor: 46.0 C
Die ID    Cores ON    Cores OFF    %
0   45   3    93.8
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    ON    OK    11.8 V    0.742 V    37.6 A (27.9 W)
1    OFF    OFF    11.8 V    0.713 V    0 A
2    ON    OK    11.8 V    0.727 V    40 A (29.1 W)
3    OFF    OFF    11.8 V    0.739 V    0 A
4    ON    OK    11.8 V    0.75 V    39.2 A (29.4 W)
5    OFF    OFF    11.8 V    0.708 V    0 A
6    OFF    OFF    11.8 V    0.729 V    0 A
7    ON    OK    11.8 V    0.749 V    39.9 A (29.9 W)
3    

Temperature sensor: 51.0 C
Die ID    Cores ON    Cores OFF    %
0   48   0    100
1   48   0    100
2   48   0    100
3   0   48    0
DC/DC ID    ON/OFF    Status    Input Voltage    Output Voltage    Output Current
0    ON    OK    11.9 V    0.703 V    38.9 A (27.3 W)
1    OFF    OFF    11.9 V    0.697 V    0 A
2    ON    OK    12 V    0.734 V    41.2 A (30.2 W)
3    OFF    OFF    11.9 V    0.733 V    0 A
4    ON    OK    12 V    0.695 V    39.7 A (27.6 W)
5    OFF    OFF    12 V    0.691 V    0 A
6    OFF    OFF    11.9 V    0.819 V    0 A
7    ON    OK    12 V    0.818 V    15.9 A (13 W)

[jelin1984]

LAST CORE HAVE 48 OFF
AND 15A

ALSO AND MY OTHER CORE HAVE NOT A LOT OF AMBER

I HAVE BOARDS WITH 8 VRMs. ALL

[jelin1984]

with firmware 0.90-0.91

the wrong board
have only 48 off

with firmware 0.92-0.95
the wrong board have ALL THE CORE OFF

why that happen???

[RoadStress]

Why are you people still using anything other than 0.95? It's the best firmware today!

[jelin1984]

because with firmware 0.95

1 chip of my jupiter is total off

i can not undresatnd why but total off

[DPoS]

because with firmware 0.95

1 chip of my jupiter is total off

i can not undresatnd why but total off

it is a shame that you have 8 VRMs and not one of the extra 4 are used to failover for the failed one.  I thought that was the point

Since you have some amps on the bad VRM the modded firmware is able to detect you have all 48 cores off.


You do have good readings on the rest.  (not many cores off)


a bad VRM can create the FPGA error loops which will cripple a board.  Sometimes the whole miner (like mine if I try .93 now)

Do you see any FPGA errors when running?  

If you get best results with .91 then stick with that


I miss the performance of .93 since it would restart cgminer if too many HW errors, but .94 seems to just let the miner slide down to bad performance.  I can start around ~440Gh with my bad miner but after a few hours it is down to ~400Gh

I will probably put a restart command in the cron for every hour or 30 minutes if that proves to be optimal.

[jelin1984]

do you beleive that is vrm that cause the problem?

what knc tell you???


why knc does not make any custom firmware for that problem?


also knc had tell me that if i will run the firmware for 1 day will fix the problem
do you have it?
run 24 hours the same firmware?

[Sitarow]

Update on 8 VRM Jupiter running firmware 0.9.5 Setup with open case with added house fan providing airflow over all components.

Stable environment conditions with an ambient room temp of 17°C / 62.6°F

24 hour results




Detailed Results Firmware v0.9.5 over 24 hours.


Compare with Results for Firmware v0.9.4 over 24 hours.



You can tell that over the course of the 24 hours firmware 0.9.5 gave more stable speeds and less hardware error %.
However firmware v0.9.4 gave more submitted shares and a WU rate of over 8200 steady.

Both firmware power usage fluctuated between 578 and 580 watts.

[viriat0]

WOW nice setup!

My friend accept donations

good luck

[Sitarow]

WOW nice setup!

My friend accept donations

good luck

Thank you.

However my motivation is to help those who have been having troubles and those that are just now getting their units.
If you wish to donate then perhaps we as a community can raise the funds so the cgminer team can get ahold of a saturn and make our lives easier

[Fishbones]

You can tell that over the course of the 24 hours firmware 0.9.5 gave more stable speeds and less hardware error %.
However firmware v0.9.4 gave more submitted shares and a WU rate of over 8200 steady.

Both firmware power usage fluctuated between 578 and 580 watts.

Thanks for the detailed analyses Sitarow.
Could you also please forward KnC Support this info and link to this thread, especially regarding the lower WU with 0.95. Seems most 4VRM and 8VRM units are experiencing this. It's costing me 0.05BTC/Block 

Keep up the good work.

Cheers.

[OC19850520]

You can tell that over the course of the 24 hours firmware 0.9.5 gave more stable speeds and less hardware error %.
However firmware v0.9.4 gave more submitted shares and a WU rate of over 8200 steady.

Both firmware power usage fluctuated between 578 and 580 watts.

So more shares is better. v0.9.4 wins, or I missing something? How much luck plays role here?

[Sitarow]

You can tell that over the course of the 24 hours firmware 0.9.5 gave more stable speeds and less hardware error %.
However firmware v0.9.4 gave more submitted shares and a WU rate of over 8200 steady.

Both firmware power usage fluctuated between 578 and 580 watts.

So more shares is better. v0.9.4 wins, or I missing something? How much luck plays role here?

Thus for 0.9.4 does win only if you can lower your ambient room temp under 21°C and know that if the temp goes up the system will throttle and even if the temp does go back down. You will have to issue a software restart in order to gain back the performance lost.

Also if you have the ability to see what the power draw is at the wall. If it is over 600 watts then you may have to contend with the reality that one or all of your ASIC boards having VRM issues.

Lowering temps seems to solve that problem.

However I will be conducting the next test by raising the ambient temp in the room to 25°C.

This is to test what the system does over the next 12 hours.

[AFox]

You can tell that over the course of the 24 hours firmware 0.9.5 gave more stable speeds and less hardware error %.
However firmware v0.9.4 gave more submitted shares and a WU rate of over 8200 steady.

Both firmware power usage fluctuated between 578 and 580 watts.

So more shares is better. v0.9.4 wins, or I missing something? How much luck plays role here?

I confirm that 0.94 is better for WU and pool hashrate. I submitted this in another topic :

Should I change pools or simply go back to firmware 0.94
Here's an average on 24 hours with firmware 0.94 and 0.95 :

--------------------	0.94	0.95
Average CGminer	270Gh/s	274Gh/s
Average Bitminter	275Gh/s	263Gh/s
Consumption	485 Watts	305 Watts
Temputure	54 & 62°C	42 & 47°C
HW	2.81%	0.79%

0.94
0.95

[Sitarow]

You can tell that over the course of the 24 hours firmware 0.9.5 gave more stable speeds and less hardware error %.
However firmware v0.9.4 gave more submitted shares and a WU rate of over 8200 steady.

Both firmware power usage fluctuated between 578 and 580 watts.

So more shares is better. v0.9.4 wins, or I missing something? How much luck plays role here?

I confirm that 0.94 is better for WU and pool hashrate. I submitted this in another topic :

Should I change pools or simply go back to firmware 0.94
Here's an average on 24 hours with firmware 0.94 and 0.95 :

--------------------	0.94	0.95
Average CGminer	270Gh/s	274Gh/s
Average Bitminter	275Gh/s	263Gh/s
Consumption	485 Watts	305 Watts
Temputure	54 & 62°C	42 & 47°C
HW	2.81%	0.79%

0.94
0.95

Thank you. The only benefit I can hope to get out of 0.9.5 is that it is able to recover from a throttled condition, or best case would be that the firmware could manage the temp increase in such a way as to not lower performance.

[r3animation]

You might want to give 0.93 a try.

I think that gives the best performance so far for Saturns.

[DPoS]

.93 is best for making money unless you have a beat up miner that cant escape FPGA error loops

.94 will let you play in the same ballpark as .93 but it lets you slide downhill if you don't watch it.
I created a script to restart cgminer on a timed loop. (they didn't install crontab)  I had it set for 30 mins but extended it to 1 1/2 hours after I cooled the chips more.

Sadly, changing my one miner to .95 degraded it but I have forced it to hash around ~420 with .94 and auto restarts.
The other miner is still ~520 running .93 solid and hasn't even restarted cgminer in days

I am starting to think I might not have a dead VRM but perhaps a whole die of 48 cores is dead so the VRM shows < 1 amp.  


 I vented enough to KNC via email and wonder if they will say anything about reducing the VRMs to 4 before actually testing the result of such a blatant bait and switch reckless move.


I am working on redesigning the case... I have a few ideas to explore.  Just about anything is better than what they slapped together for jupiters


and here's my pain summed up in a picture:

[mininganon]

An Update on my Jupiter: After taking some advice on the KNC forum I updated to .95, rebooted then did a "hard reset" as outlined in the KNC manual. I saw .95 wasn't working out and switched to .94. I'm now stable at .94 with a cool room where before the hard reset it would drop ghs to unacceptable levels. I'm getting this on Eligius.st right now
3 hours   473.45 Gh/s   1190528
22.5 minutes   475.64 Gh/s   149504
256 seconds   551.80 Gh/s   32890
128 seconds   523.45 Gh/s   15600

My only issue now is this module marked with a "3" sticker on it. My other three modules have "1" and are fine.

Temperature sensor: 31.5 C
Die ID    Cores ON    Cores OFF    %
0   48   0    100
1   0   48    0
2   48   0    100
3   48   0    100
DC/DC ID    ON/OFF    Status    Input Voltage    Output Voltage    Output Current
0    ON    OK    11.8 V    0.744 V    34.9 A (26 W)
1    OFF    OFF    11.9 V    0.712 V    0 A
2    ON    OK    11.9 V    0.815 V    15.1 A (12.3 W)
3    OFF    OFF    11.9 V    0.825 V    0 A
4    ON    OK    12 V    0.745 V    38.6 A (28.8 W)
5    OFF    OFF    11.9 V    0.695 V    0 A
6    OFF    OFF    11.8 V    0.739 V    0 A
7    ON    OK    11.9 V    0.732 V    40.2 A (29.4 W)

[jelin1984]

exactly the same and mine

the faulty core
have stickers number 3

but i can run it only with 0.91

[sickpig]

An Update on my Jupiter: After taking some advice on the KNC forum I updated to .95, rebooted then did a "hard reset" as outlined in the KNC manual. I saw .95 wasn't working out and switched to .94. I'm now stable at .94 with a cool room where before the hard reset it would drop ghs to unacceptable levels. I'm getting this on Eligius.st right now
3 hours   473.45 Gh/s   1190528
22.5 minutes   475.64 Gh/s   149504
256 seconds   551.80 Gh/s   32890
128 seconds   523.45 Gh/s   15600

My only issue now is this module marked with a "3" sticker on it. My other three modules have "1" and are fine.

Temperature sensor: 31.5 C
Die ID    Cores ON    Cores OFF    %
0   48   0    100
1   0   48    0
2   48   0    100
3   48   0    100
DC/DC ID    ON/OFF    Status    Input Voltage    Output Voltage    Output Current
0    ON    OK    11.8 V    0.744 V    34.9 A (26 W)
1    OFF    OFF    11.9 V    0.712 V    0 A
2    ON    OK    11.9 V    0.815 V    15.1 A (12.3 W)
3    OFF    OFF    11.9 V    0.825 V    0 A
4    ON    OK    12 V    0.745 V    38.6 A (28.8 W)
5    OFF    OFF    11.9 V    0.695 V    0 A
6    OFF    OFF    11.8 V    0.739 V    0 A
7    ON    OK    11.9 V    0.732 V    40.2 A (29.4 W)

Same here. 3 board with a '1' sticker on top. The fourth is marked with a '3'.
Bertmod show me that one my asic slot has a die with all 48 cores disabled.
So i thought that knc did this  because they didnt have enough well functioning
chips for everybody.

Last piece of info. All my board have 8 VRM. I'm not able to run .92 e .93.
.90 run only for the first few mins, then upgrade to .91. Both .94 and .95 with almoste the same perf