Swedish ASIC miner company kncminer.com

[edgar]

As someone who had a shoddy boxed unit full of rainwater arrive I've been quite pleased with the Jupiter...it runs on .95 at 555 gh and the highest core temp I have is
40c one is 28c the other 2 32c each....

so to say the least I've not upgraded ANYTHING but take the top off and have 2 small walmart fans pointing at the naked miner..that's it...been too scared I'd muck it up...to even think about anything else at least till I saw .97 as the recommended update

anyway if I do go with the new .97 upgrade this weekend when I have time to panic if it goes wrong....how hard is it to backtrack...say to .95 or 96? not that big a deal or what?

if it is painless going backwards as well as forward I'll take the chance

awaiting better minds then myself to weigh in on this


I'm also on slush an my ave hash there of long term is about 525gh even....

i'm not at work so I don't have hw nor wu rates available I think my error rate last I looked was 5 percent

anyway whatcha think

hey I appreciate the help ..I still think this so called out of control ahem "hobby" would be better run as co-op with some guys over beers...its lonely out in the sticks when you bump the table and your miner goes off and you scream like an 8yr old girl in terror!!! (luckily re-boot pops it right back up on restart)....such things are better shared while telling bfl horror stories over a campfire and a beer imho

anyway thanks for the advice you guys are pretty much the only source in town

Searing

If you are happy with its performance at .95 i would suggest you leave it as is - i was happy but wanted more and regret fkn with it now.



Also, try to think what you want to say and post it in 1 msg, if you need to add more use the edit button - multiple posts again & again gets quite annoying and i personally begin ignoring them..

@rawbit - did all that no joy - flashed back to .97 and everything is back to normal so far...

PuTTy & KnC Gui both show my avg as 539 but BTCG varies constantly from 480 - 569. Not sure why, perhaps the flushing still?

[1713912851]

1713912851

[1713912851]

1713912851

[PilotCurly99]

Anyone seen an issue with a Jupiter that is hashing at half speed?  Appears by temperature, that two of the cores are not running at all. 
I;ve tried moving cables to different slots, hard resets, and different versions of firmware.  Every time, half the speed and only two cores show an increase in temperature.

Any ideas?  Just got it today and very disappointed.  My first one works perfectly, so this is very surprising.

[bobsag3]

Also have a strange unit.... Worked fine for ~8 hours, then started randomly shutting off.
Tried different PSU's, ethernet cables, internal connections checked, firmware from .95 to .97 and back, nothing. Unit just hashes away @ ~500ghash with 1 ASIC being much hotter than the others. Finally I tried unplugged that ASIC and boom- rock solid stable and hashing away at 410ghash. Do I have a dead asic or is this something I can resurrect?

[FoolPartedWithMoney]

Anyone willing to run their unit with chip fans removed along with me? 'Cause it's a bit nerve-racking, and it will make me feel a bit easier.

[trepex]

So it's official that 2 Saturns > 1 Jupiter (not by much, but still).

From a logic point of view: Of course!
Assuming that the ASICs perform at the same speed for both models under the same environment.

BUT
- 2x Saturn have better cooling (you can fix this with manually adding extra cooling/fans)
- The flushwork and maybe some other CPU/FPGA actions rely on the BBB speed. For a Jupiter the BBB has twice the
   amount of work to do.

On the other side:
- A Jupiter is more power efficient.
- A Jupiter had been cheaper than 2x Saturn

[Puppet]

From a logic point of view: Of course!
Assuming that the ASICs perform at the same speed for both models under the same environment.

BUT
- 2x Saturn have better cooling (you can fix this with manually adding extra cooling/fans)
- The flushwork and maybe some other CPU/FPGA actions rely on the BBB speed. For a Jupiter the BBB has twice the
   amount of work to do.

It really shouldnt have to. A BeagleBone is considerably more powerful than a raspberry pi and I would be surprised if it struggled running 4 or many more modules. If its true the bottleneck is there, it has got to be fixable with a less crappy firmware. And if it really is too slow, it would make sense to just buy a second BBB or Pi, you'd get better performance per $.

Likewise for cooling, if the bottleneck is there, then KnC shipped too slow or not enough fans. A pair of exhaust fans costs like $2, why wouldnt they include that for jupiters then? You are paying what, $10 per GH? If more fans would add just 1% more hashrate, its a massive win.

[ElGabo]

I've screwed up 0.97. Hashrate down from 550 to 410. HW from 3,5 to 10%.

It was runing for 12 hours. Now back to 0.96.

 

[kitcut]

I am sorry suddenly, mate.

I want to ask someone.

Is it possible to switch the order to hosting service?

Is there the person who was able to do it?

[operador]

An update on the hosting front, my Jupiter has been completly offline for 14 hours, I am a day 1 delivery customer and I have been hashing for a long time now.
No news and support told me about how I should click save only once when the problem is clearly not related to the WEB UI!!!!
I know of at least 4 other people in the same boat right now.

Not sure what's left to do, hoping this will get reserved shortly but man, I am losing money every day with this thing....

[pgbit]

In case this helps anyone, I just upgraded my Mercury from 0.95 to 0.97, and got a cgminer failed to start error:
"Running (Connect to CGMiner API failed)"
I was already connected to Eligius at the time and a restarts of cgminer / reboot (three times) with browser cache clearance did not help.
To get it to work, on the web browser Miner interface, I checked the CGMiner API Enabled checkbox, saved changes, and restarted Cgminer then the machine and then everything updated just fine. After a few mins, the hash rate has improved from 120 GHs to just above 140Ghs...

[naRky]

I've screwed up 0.97. Hashrate down from 550 to 410. HW from 3,5 to 10%.

It was runing for 12 hours. Now back to 0.96.

 

as i see i will not upgrade to 0.97, my jupiter has 540GH/s average and 1.9% HW errors with 0.95

[ElGabo]

If something not broken.....

 

[FeedbackLoop]

0.97 also shows a continuous drop of hashrate for me from 500 to 470 (3 hour average eligius).

[Xialla]

.97: after couple of hours my hashrate dropped from 550 to 450 without any reason but reboot solve it. now considering downgrade to .96 - it runs at 540 for long days without this issue..

[realtime]

Edit: My bad wasn't two only one....  Sorry!

[btc_uzr]

I don't know if anybody has checked but can someone check...i'm VERY confused....

I'm show only 627W for 2x Jupiters.....Running @ 552 and 548 Hash....313W per machine?! how did they drop 200W?!

0.97 ftw?!

  ..313W per machine for real ?!

[sickpig]

I don't know if anybody has checked but can someone check...i'm VERY confused....

I'm show only 627W for 2x Jupiters.....Running @ 552 and 548 Hash....313W per machine?! how did they drop 200W?!

0.97 ftw?!

How did you measure? bertmod or at the wall?

[realtime]

Edit...*hang head in shame*.....it was only one.....triple checked my wiring...my bad

[sickpig]

At the wall, uploading video proof now, and photo of meter.

Confirming i'm not crazy 623W (240V) @ the wall with a total hash rate of 1.1TH~!

CRAZY!

That is for 2 Jupiters!

very very good !


edit1: if you don't mind could please use bertmod (asic_status.pl) to get the VMR DC/DC power consumption ?

[naRky]

At the wall, uploading video proof now, and photo of meter.

Confirming i'm not crazy 623W (240V) @ the wall with a total hash rate of 1.1TH~!

CRAZY!

That is for 2 Jupiters!

FW??