Topic: KnCMiner Jupiter Miner First Impressions - page 18. (Read 65993 times)

I upgraded to .95 from .93 for one miner that ran 770 watts and 520 Gh (rated at the pool, not cgminer since cgminer doesnt pay you anything)

the wattage dropped to 460w and the mining dropped to 490 Gh and temps dropped about ~10C on each slot

so I shed 300 watts by giving up 30Gh,  I guess that deserves to stay for a while and let everything cool down since it is using a 850 watt psu


I added bertmod and found something interesting.   One of my dies is terrible.  running 50%. (24 cores off)  most others have one or two cores off but I do have one slot that all 4 are perfect



So .93 pushes more out of my non-perfect miner as long as I kept it cool enough to function.  And was able to run it 770watts on a 850psu though..  not something to try long term.


so I will leave this one at .95 and keep the other one on .93 since it has a 1000w psu for now and at least has one board with 8 VRMs


also, to toss a conspiracy theory out there, the one board that is perfect has sticker #1.  the other three all have sticker #2 and have atleast a few cores off and one die that 50% are off

Sweet!  I just upgraded to .95 and my wattage from the wall dropped from 900 to 570.  I'd say thats a problem solved.  Good work KNC  

Performance wise, I was at around 450ghash in CGMiner on the .94 firmware, now I'm seeing an average of 520 in cgminer...and rising  

Too soon to tell on stability as I upgraded only 20 minutes ago.

MINE CORE work only with 0.91 and 0.92 firware
with any other is total dead

i can understand why???

yeah, same here. WU is lower..don't know if it is good or bad mark.) less HW errors and ~545Gh/s average hashrate. Will see tomorrow.

Sitarow,  you might not see much of a performance jump with .95 since your Jupiters all have 8 VRMs right?

I may upgrade one of mine that has all boards with 4 VRMs since it is really pushing my 850psu at 770 watts

Not sure if it matters but my WU/m is down with the new update. From 40xx WU/m to 37xx WU/m.

Pool hashrate is around the same and HW errors have dropped alot.

Here is the results after at least 10 minutes.

I will wait as well.  .93 is very good at bringing hash rate back up quickly

yeah it is a miracle I was able to cool my jupiter down enough that had all 4 boards with 4 VRMs using a 850 PSU and hash stable ~520  (770 watts)

but orama says I have horse ass for brains thinking cooling mattered

First results Firmware v0.9.5 power draw is 582 watts at the wall.

Case cover off and house fan put on top of unit.

This does nothing to help the poorly designed work restarts.

Firmware v0.9.5 is necessary for those with 4 VRM on the asic modul. It is the only reasonable option at this time if you wish to use all 4 modules on a 850 watt power supply.


source

http://forum.kncminer.com/forum/main-category/hardware/6746-asic-bords-with-4-vrms-bug-cool-properly-until-fw-fix-arrives

That may be true. However this exercise is an attempt to get everyone up to speed including the manufacturer.

Hopefully it does work.

This is the last reading I am going to take using firmware v0.9.4

From the looks of it I would say that given time it would eventually get up to the average rate of 544 GH/s.

Let us see how 0.9.5 works.

link?   a lot of us rather not roll the dice since they do not do any testing like Sitarow does.

They offer no metrics, nothing, but just say a few things.   Same reason they took off the 4 VRMs, since someone 'thought' it was better to not use them.

I may have killed it before when the system had stalled.


  PID USER       VSZ STAT COMMAND
    1 root      1652 S    init [5]
    2 root         0 SW   [kthreadd]
    3 root         0 SW   [ksoftirqd/0]
    4 root         0 SW   [kworker/0:0]
    5 root         0 SW<  [kworker/0:0H]
    7 root         0 SW<  [kworker/u:0H]
    8 root         0 SW   [migration/0]
    9 root         0 SW   [rcu_bh]
   10 root         0 SW   [rcu_sched]
   11 root         0 SW   [watchdog/0]
   12 root         0 SW<  [khelper]
   13 root         0 SW   [kdevtmpfs]
   14 root         0 SW<  [netns]
   15 root         0 SW   [kworker/0:1]
   16 root         0 SW   [bdi-default]
   17 root         0 SW<  [kintegrityd]
   18 root         0 SW<  [kblockd]
   19 root         0 SW   [khubd]
   20 root         0 SW   [irq/86-44e0b000]
   21 root         0 SW   [kworker/u:1]
   24 root         0 SW   [irq/23-tps65217]
   27 root         0 SW   [irq/46-4819c000]
   36 root         0 SW   [irq/87-4802a000]
   39 root         0 SW<  [rpciod]
   41 root         0 SW   [khungtaskd]
   42 root         0 SW   [kswapd0]
   43 root         0 SW   [fsnotify_mark]
   44 root         0 SW<  [nfsiod]
   45 root         0 SW<  [crypto]
   48 root         0 SW<  [pencrypt]
   49 root         0 SW<  [pdecrypt]
   56 root         0 SW<  [OMAP UART0]
   58 root         0 DW   [spi1]
   61 root         0 SW   [spi2]
   64 root         0 SW<  [kpsmoused]
   65 root         0 SW   [irq/150-mmc0]
   77 root         0 SW<  [deferwq]
   78 root         0 SW   [kworker/u:2]
   79 root         0 SW   [mmcqd/1]
   80 root         0 SW   [mmcqd/1boot0]
   81 root         0 SW   [mmcqd/1boot1]
  114 root         0 SW<  [kworker/0:1H]
  321 root      2148 S    udhcpc -b -x hostname Jupiter-20E eth0
  335 root      2272 S    /usr/sbin/dropbear -r /config/dropbear_rsa_host_key
  338 root      3272 S    /usr/bin/ntpd -p /var/run/ntp.pid -g
  346 avahi     2820 S    avahi-daemon: running [Jupiter-20E.local]
  347 avahi     2728 S    avahi-daemon: chroot helper
  353 root      3108 S    /usr/sbin/lighttpd -f /etc/lighttpd.conf
 1144 root      2016 S    sleep 60
 1145 root      2328 R    ps
 1646 root      2772 S    /usr/sbin/dropbear -r /config/dropbear_rsa_host_key
 1677 root      2328 S    -sh
 1678 root      2616 S    screen -dr
 1704 root      2616 S    {screen} SCREEN -S cgminer -t cgminer -m -d /usr/bin
 1705 root      112m S    /usr/bin/cgminer --default-config /config/cgminer.co
 1708 root      1496 S    /usr/bin/monitor-pwbtn /usr/bin/factory_config_reset
 1725 root      1908 S    /sbin/getty 115200 ttyO0
 1726 root      1908 S    /sbin/getty 38400 tty1
 1727 root      2148 S    {monitordcdc} /bin/sh /sbin/monitordcdc
15845 root      2772 S    /usr/sbin/dropbear -r /config/dropbear_rsa_host_key
16363 root      2328 S    -sh
31784 root      2944 S    /usr/sbin/dropbear -r /config/dropbear_rsa_host_key
31838 root      2328 S    -sh
32463 root      2328 S    sh
32484 root      2328 S    sh -l
root@Jupiter-20E:~# ps | grep asic_status | grep -v grep
root@Jupiter-20E:~#

yep thanks and also the first feedback seem encouraging

It does not want to die!

  766 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  768 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  770 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  772 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  774 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  776 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  778 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  780 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  782 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  784 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  786 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  788 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  790 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  792 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  794 root      2152 S    grep asic_status
root@Jupiter-20E:~# ps | grep asic_status
  796 root      2152 S    grep asic_status
root@Jupiter-20E:~#


Edit: Okay will do.

just to be safe, once logged in via ssh, check if asic_status.pl is still running:


if yes just stop it using  kill command,


the pid is the first number you'll see in the output of the first command that I typed. otherwihse you can use pidof.

i think winzip will deal with tar.gz quite nicely.

from the same place where you've downloaded putty you can download pscp, a command line to run from command prompt with the same exact syntax.