Topic: Swedish ASIC miner company kncminer.com - page 1406. (Read 3050074 times)

thanks for that detail - some were saying they saw differences on the amount of paste.   I didn't know it was automated but it is a sensitive part of cooling so all eyes on that when temps are all over the place

In the first photo the upper left hashing board is it just a camera trick or is one of the heatsink mounting screws missing?

Beautiful.  You're a lucky guy, KnC miners and you live in Hawaii.

ps aux?

my miners after delivery:

What are you on about? each chip is pasted prior to assembly an the manufacturing facility. They have a special metal grid that lies flush on top of the chip and the paste is spread thinly over.

Okay, back from working outside.  The poster was correct, the # is the command prompt and if you add a user, when you login as the user the command prompt will be $ so if you see the # prompt you have root privileges, if you have the $ you have user privileges.

did you run #cgminer --help or #cgminer --help | more  

if you are connecting to eligius then the command will be #cgminer -o stratum+tcp://mining.eligius.st:3334 -u (here put your bitcoin address to which the bitcoins will be sent) -p (anypassword)

From what height prey tell??!

Also why would a box be dropped clearly on top of the side of the unit when there are obvious arrows pointing in the direction of 'up'.

wait what about paste???  just some pre-paste on these?   that would explain why some modules are hotter than others beside the other 10 reasons we have talked about

Dont be silly. Its clear some other boxes were dropped on top of that one. Sturdier boxes I might add, KnC really should use better packaging.

mine as well, even after I begged them to ship the fans and heatsinks unassembled





one or two out of 8 ain't bad for these guys!!!!

Holy shit! There's actually a joke that goes along with that, which appears a little insenstive now. And as a result a miner was actually packed in that box by mistake. I actually drew that on there myself, but I swear that left with UPS in mint condition;



Thanks for posting that, that actually appears to have been damaged purposely and maliciously, and as far as i'm concerned should burn UPS's delivery contract.

And no Xialla, get your facts straight KnC is NOT ignoring complaints, they've ordered a lot stronger boxes, and are awaiting delivery of them, as mentioned earlier.

KNC is ignoring complaints about packaging. Official forum is flooded by photos of disassemble coolers after delivery and holes in cartoon boxes. No one cares and week after first complain, KNC don't change anything in packaging.

 

"But surely if we hand write "fragile" in marker pen, they'll know it's a piece of delicate electronics worth $8000 and handle it carefully?"

Bitcoinorama, First of all: THANK YOU, for your great reporting!

But now to the bare facts:

I have received my Jupiter today and tested it thoroughly (0.91, 0.93 and 0.94) and this on all 4 ASIC-boards as one Jupiter-system, as well as each ASIC-board separately (only connecting one board at one time). And found huge differences in the boards.

Actually only one of my Jupiter-boards is working very well, one is running more or less OK (has some Core disables), but 2 are just bad. Together they make only 390 Gh/s (the good one is 135 Gh).

Be aware that this Jupiter is cooled very well (in a Server-room with 18 C, without the case-top, and even an extra Fan blowing 18 C cool air into the Jupiter-box).  So hardware-errors due to heat problems are impossible.  Still the bad boards have about 30% of bad cores…. Reducing the optimum hashrate from 140 G/s per ASIC-board to 100 or less.

Sure the FW versions 0.91 to 0.94 are trying to “stabilize” the system, but actually it’s only disabling the faulty cores per ASIC. This disabling results in less reporting of HW-errors indeed, and so reduces the Hashrate performance as well (by disabling 30% of the cores). This disabling is also noticeable if you look at the Temperatures of the ASICS (and their Watt use). See this list (sorry, I don't know how to insert a screenshot into this post):

ASIC slot #1: 57 C
ASIC slot #2: 36 C
ASIC slot #3: 38 C
ASIC slot #4: 41 C

the one with 57 C runs very well (on its own about 135 Gh).

So looking to all other posts and reports about Jupiters running 550 Gh and others like me running only 391 or lower: You simply are lucky if you received a Jupiter with 4 good ASIC-boards (then 4 times 135 is about 540). If you have one or more not so well manufactured ASIC-boards you end up (much) lower.  I’m afraid (please ask Sam for me) that software updates will NOT solve this thing. We need to replace the hardware: more boards per Jupiter (tested at KnC before shipping) that are capable of running 140 G without Hardware-errors.
 

when you typed screen ~r it replied [screen is terminating]

I have no idea what that meant but you could of wacked something.  Might be good to just reboot miner (if you dont have a corsair psu lulz) as long as you haven't touched anything else

if you already messed with files then you have to retrace your steps

Thanks guys its all sorted and I'm now up and running if anyone else needs help then I suggest this guy---->  https://bitcointalksearch.org/user/trepex-102694

He was very helpful and guided me every step of the way also helped me with the ssl for cgminer Thanks

Potato tech maybe.  (Just kidding).

Uh, really?

Google Micron Technology.

Boise/Treasure Valley area is a high tech mecca like Silicon Valley. There was, of course, no internet because there was no internet anywhere But we had tech. That was why it was so hard to get an apartment. Zilog, HP, Micron, and several startups were hiring like mad. Boise doubled in size in the early 90's on a high tech boom.

Great!
By tabs they mean keys.  pff.