Swedish ASIC miner company kncminer.com

[Phoenix1969]

Unix/Linux is also known to parallel process rather slow...

wow Phoenix1969. this is bold.
citation needed at least  

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=5&cad=rja&ved=0CFIQFjAE&url=https%3A%2F%2Fforums.oracle.com%2Fthread%2F2546176&ei=SPd5Uu_iOsmwiQKW7YCYBQ&usg=AFQjCNGuoC9AUd5HXvRhoRsJvm7A1PncOg&sig2=-ZH3Y3uynYTbhEXRFUSpBw&bvm=bv.55980276,d.cGE
I could post a hundred references, but this isn't the place... 

[1714066835]

1714066835

[timmmers]

Two firmware updates for me so far. Didn't do that, bot even half of it. Not had a problem

1. DL the update
2. On the KNC miner interface click update and select that download and apply it.
3. Click restart cg miner in the KNC interface.

Nothng else at all. Checked it after in Putty, but all was fine. lol

So after you upload and apply the firmware, it returns to a screen that says restart miner, is that the link you clicked or did you go back to the home screen, then go the the "Mining" and click on "Restart CGMiner" on that page?

Tbe last one, Mining screen >> Restart CGMiner.

[Phoenix1969]

Two firmware updates for me so far. Didn't do that, bot even half of it. Not had a problem

1. DL the update
2. On the KNC miner interface click update and select that download and apply it.
3. Click restart cg miner in the KNC interface.

Nothng else at all. Checked it after in Putty, but all was fine. lol

So after you upload and apply the firmware, it returns to a screen that says restart miner, is that the link you clicked or did you go back to the home screen, then go the the "Mining" and click on "Restart CGMiner" on that page?

Tbe last one, Mining screen >> Restart CGMiner.

no wonder

after you "apply" a message will appear with reboot in blu...hit that!
alternatively, you can power cycle the miner after apply.. but no need.
firmware wont apply without a reboot, restarting cgmner isnt enough...

[sickpig]

Unix/Linux is also known to parallel process rather slow...

wow Phoenix1969. this is bold.
citation needed at least  

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=5&cad=rja&ved=0CFIQFjAE&url=https%3A%2F%2Fforums.oracle.com%2Fthread%2F2546176&ei=SPd5Uu_iOsmwiQKW7YCYBQ&usg=AFQjCNGuoC9AUd5HXvRhoRsJvm7A1PncOg&sig2=-ZH3Y3uynYTbhEXRFUSpBw&bvm=bv.55980276,d.cGE
I could post a hundred references, but this isn't the place... 

I can assure you that "Solaris 11.1 SRU 3.4.1." is quite different from modern linuxes regarding a lot of things (scheduler/memory allocation/IO etc etc etc). So I wouldn't have inferred  that since Solaris version XYZ has a problem then also Linux ver KYZ will behave in the same way in the same conditions. 

 

[Bitcoinorama]

https://www.kncminer.com/news/news-63


"Unbeknown to the community we are currently developing the next generation product in 20nm/16nm process with Alchip "

They have time to post such news, but don't have the time to participate in their own forums and help people out on their devilish products? They aren't thinking people are going to buy from them again, do they? Is making an announcement like that just a start of mind controlling techniques in the marketing world? Also, if you tried to visit that link www.ast.com that is IN their news publication it doesn't even work.

1. The press release is from Alchip, not KnC. It's hardly mind control.
2. There just aren't enough bodies currently to maintain the forum ontop of all else. If you honestly think the reason is they cannot be bothered that's either naivety or ignorance. They have a dedicated customer support, the forum isn't a priority in my mind. It's one of the things they want me to work on and I'm flat refusing to be honest. I'm a bit disenchanted as to how I was treated here, entire threads abusing me, I'm not keen on more. I'd rather customer service focused on individual issues, and when/if my other work load dies down I'll consider it, but everytime I post I get multiple PMs.
3. Personally my own opinion i'e always followed and enjoyed the banter in Bitcointalk.
4. Thanks for letting me know about AST's deadlink, again it's Alchip 'a release. You can see it was posted y'day online. I presume they either added the incorrect link, or the site is down.


Edit: Here's the OG; http://www.digitimes.com/news/a20131105PR202.html

And AST's correct link is; www.ast.co.il I'll get it fixed

[Tigggger]

Well after 5 days of nonstop stable mining my Jupiter just stopped hashing. The web interface said CGMiner was down and when SSHing in there was no screen session running. Restarting the miner via the web interface did nothing but a reboot has it back in action. Only 10 minutes of downtime but guess I still need to keep my eyes peeled for offline email warnings. Hopefully they'll eventually get it as stable as my Avalons which I have to say are 100% stable even when overclocked. Going to look at the logs when I have time tomorrow...

I've been noticing this since 0.98+ and the devs are aware of the issue and are working on it.

I lost 7 hours of mining the other day as it happened at 2am and didn't catch it until I woke up.

I have now set up an alert on my phone to catch the idle notification emails and play an siren to wake me up.

[Mota]

https://www.kncminer.com/news/news-63


"Unbeknown to the community we are currently developing the next generation product in 20nm/16nm process with Alchip "

They have time to post such news, but don't have the time to participate in their own forums and help people out on their devilish products? They aren't thinking people are going to buy from them again, do they? Is making an announcement like that just a start of mind controlling techniques in the marketing world? Also, if you tried to visit that link www.ast.com that is IN their news publication it doesn't even work.

1. The press release is from Alchip, not KnC. It's hardly mind control.
2. There just aren't enough bodies currently to maintain the forum ontop of all else. If you honestly think the reason is they cannot be bothered that's either naivety or ignorance. They have a dedicated customer support, the forum isn't a priority in my mind. It's one of the things they want me to work on and I'm flat refusing to be honest. I'm a bit disenchanted as to how I was treated here, entire threads abusing me, I'm not keen on more. I'd rather customer service focused on individual issues, and when/if my other work load dies down I'll consider it, but everytime I post I get multiple PMs.
3. Personally my own opinion i'e always followed and enjoyed the banter in Bitcointalk.
4. Thanks for letting me know about AST's deadlink, again it's Alchip 'a release. You can see it was posted y'day online. I presume they either added the incorrect link, or the site is down.


Edit: Here's the OG; http://www.digitimes.com/news/a20131105PR202.html

And AST's correct link is; www.ast.co.il I'll get it fixed

I am sorry that you took some heat here. We all know that you meant well and that you were pretty much a neutral party until you joined knc. But you could have known there would be a shitstorm coming your way the moment they messed up their delivery for some people while others got their orders on time. Again, not your fault, I know. It was just very very VERY badly handled by KNC.
Also, there is no dedicated customer support. There are 4 people working customer support for a multi million dolalr company. But at least those 4 are halfway competent and know their jobs.

[semaster]

over thousand pages in thread.
Seems like we need new topic "KnCminer November orders\shipping"

[Biffa]

It seems no one is interested in external binaries any more, but just in case they are, here is a binary for the knc devices based off the latest cgminer 3.7.0 code. There are no performance advantages for knc devices but there are networking and work scheduling fixes for people that have had issues (eg difficulty connecting to eligius or system stopping work entirely). As per previously, you'll have to ssh into and copy the binary to your device to run it.
http://ck.kolivas.org/apps/cgminer/kncminer/cgminer

Bugfix update to 3.7.1. Same download location.

Some beer money sent.

Everyone should send a few dollars to this man.

[lemonte]

https://www.kncminer.com/news/news-63


"Unbeknown to the community we are currently developing the next generation product in 20nm/16nm process with Alchip "

They have time to post such news, but don't have the time to participate in their own forums and help people out on their devilish products? They aren't thinking people are going to buy from them again, do they? Is making an announcement like that just a start of mind controlling techniques in the marketing world? Also, if you tried to visit that link www.ast.com that is IN their news publication it doesn't even work.

1. The press release is from Alchip, not KnC. It's hardly mind control.
2. There just aren't enough bodies currently to maintain the forum ontop of all else. If you honestly think the reason is they cannot be bothered that's either naivety or ignorance. They have a dedicated customer support, the forum isn't a priority in my mind. It's one of the things they want me to work on and I'm flat refusing to be honest. I'm a bit disenchanted as to how I was treated here, entire threads abusing me, I'm not keen on more. I'd rather customer service focused on individual issues, and when/if my other work load dies down I'll consider it, but everytime I post I get multiple PMs.
3. Personally my own opinion i'e always followed and enjoyed the banter in Bitcointalk.
4. Thanks for letting me know about AST's deadlink, again it's Alchip 'a release. You can see it was posted y'day online. I presume they either added the incorrect link, or the site is down.


Edit: Here's the OG; http://www.digitimes.com/news/a20131105PR202.html

And AST's correct link is; www.ast.co.il I'll get it fixed

Any news on the reseller portal O'rama? I know it probably isn't too high on priorities right now, but would like to see how I did.

[Biomech]

Thanks for sharing your experience. Really value it.

Aren't you? At least in part?

No. Not even in part. For that I use a day job that fulfils that need completely (fortunately) and I think that's the case with the majority of the people in this thread (not saying that mining full time is not OK but then you will be competing with hobbyists/small investors in a way not seen in other investments, which is beneficial to Bitcoin like I discussed in an earlier post). You can start drawing future scenarios for my person or define "food" and "debt" in such a way that you can say "at least in part" but then there's very little meaning in the question: "what are you mining for?".

Short Form. I think BTC will continue to rise against traditional fiat, and I think of it more as a store of value than a daily currency.

I prefer to think of it as "mostly currency" with properties that make it hold value better in it's money aspect than most FIAT currencies, but I really don't think anyone wants us to debate that here and I'm not that interested in definitions.

Cheers!
 

Fair enough, this probably isn't the place. Right now my time is so limited that I can't pursue this in great detail, but if you'd be interested in debating it some other time and in a new thread, I would be up for that. In reading my original post, it came off as an attack. In reality, I just hadn't had my coffee yet I find a lot of your stated opinions to be similar to mine, and I like your posts. I think it would be an engaging debate.

Anyway, I'll leave it at that for now, except to say that economics is a rather serious hobby of mine. That is what initially attracted me to bitcoin, it's a novel approach to money. Not exactly fiat, not tangible, but with elements of both. It seems a worthy experiment from many different angles.

[Bitcoinorama]

https://www.kncminer.com/news/news-63


"Unbeknown to the community we are currently developing the next generation product in 20nm/16nm process with Alchip "

They have time to post such news, but don't have the time to participate in their own forums and help people out on their devilish products? They aren't thinking people are going to buy from them again, do they? Is making an announcement like that just a start of mind controlling techniques in the marketing world? Also, if you tried to visit that link www.ast.com that is IN their news publication it doesn't even work.

1. The press release is from Alchip, not KnC. It's hardly mind control.
2. There just aren't enough bodies currently to maintain the forum ontop of all else. If you honestly think the reason is they cannot be bothered that's either naivety or ignorance. They have a dedicated customer support, the forum isn't a priority in my mind. It's one of the things they want me to work on and I'm flat refusing to be honest. I'm a bit disenchanted as to how I was treated here, entire threads abusing me, I'm not keen on more. I'd rather customer service focused on individual issues, and when/if my other work load dies down I'll consider it, but everytime I post I get multiple PMs.
3. Personally my own opinion i'e always followed and enjoyed the banter in Bitcointalk.
4. Thanks for letting me know about AST's deadlink, again it's Alchip 'a release. You can see it was posted y'day online. I presume they either added the incorrect link, or the site is down.


Edit: Here's the OG; http://www.digitimes.com/news/a20131105PR202.html

And AST's correct link is; www.ast.co.il I'll get it fixed

Any news on the reseller portal O'rama? I know it probably isn't too high on priorities right now, but would like to see how I did.

Sorry I meant to get back to you on that as you asked a while back. I did ask and Sam said right now there's a lot more that require prioritization, so re-seller portal is way down that list i'm afraid. They haven't really concentrated on marketing at all, just engineering, hence the reason with respect to stuff like adwords it's dominated by companies that aren't shipping anything, or haven't yet an tangible products. It's also why you haven't seen multiple press releases pertaining to every component they bulk ordered!

They are hiring more people, but you all have to understand there's risks as how the chips would perform that were unknowns until the chip is in hand. The other issue is hiring people that can work under this kind of pressure, speed and can read from the same hymnsheet w.r.t. Bitcoin. If you hire someone who isn't they slow you down, not maintain the momentum, or even increase it. The beauty of who they do have here is that they are very competent and are excellent in their role and as part of a team. There's a dedicated RMA team now, so that's taken a lot off my shoulders as I was spending a lot of time on that. They have engineers being interviewed all the time, a couple are starting next week, and more customer support are starting over the next couple of weeks.

[markm]

Yep it was out of memory, dmesg shows:

Code:

[292489.141503] Out of memory: Kill process 3622 (cgminer) score 188 or sacrifice child
[292489.149604] Killed process 3622 (cgminer) total-vm:189152kB, anon-rss:109744kB, file-rss:1652kB

-MarkM-

[sickpig]

Yep it was out of memory, dmesg shows:

Code:

[292489.141503] Out of memory: Kill process 3622 (cgminer) score 188 or sacrifice child
[292489.149604] Killed process 3622 (cgminer) total-vm:189152kB, anon-rss:109744kB, file-rss:1652kB

-MarkM-

try to contact ckolivas and report what happened in the meantime switch back to a working version of cgminer.

[markm]

Yep it was out of memory, dmesg shows:

Code:

[292489.141503] Out of memory: Kill process 3622 (cgminer) score 188 or sacrifice child
[292489.149604] Killed process 3622 (cgminer) total-vm:189152kB, anon-rss:109744kB, file-rss:1652kB

-MarkM-

try to contact ckolivas and report what happened in the meantime switch back to a working version of cgminer.

I grabbed the ckolivas one because it was said to possibly fix cases where mining stops completely. which is what I had been seeing from the official release.

I hoped that this problem of mining being stopped completely according to the web page, which has turned out to be due to an out of memory condition, was the mining stopping completely problem that the wget grabbed version hoped to fix.

In other words, I didn't have a working version in the sense of one that did not mysteriously die after a day or two or few and hoped the wget would get me one that would not thusly die.

I am trying the latest wget with --lowmem now hopefully that won't die...

But it seems to be settling at 278 GHashes whereas I had been getting usually 283, so maybe putting another stick of RAM or larger stick of RAM in the damn beaglebone might be better if --lowmem loses hashing power. I wonder if the damn beaglebones can handle more RAM? Maybe KnC skimped on RAM too quickly without considering that software tends to get bigger all the time...

Or maybe I should kill the stupid webserver, if I can ssh what for do I need a webserver on the thing anyway? Hmm...

Also, I looked at the file I had used wget to get with "file" command on an X86_64 , and that claims its an ARM executable that is not stripped. Wouldn't it maybe use less RAM if it were stripped? Does it make much sense to use a not-stripped executable on an "imbedded device" type thing like this where RAM is at a premium and no debugger is installed? Any chance of a stripped version, ckolivas? (My X86_64 strip doesn't even recognise the thing as an executable and the miner does not have strip installed...)

-MarkM-

[Biffa]

What version are you using? I'm using 3.7.2 over 0.96 firmware.

3.7.0 had some bugs, and 3.7.1 was broken.

cgminer is using about 23% of memory on the device. Looks like the device has 512MB of mem.

My machine has been running for 24 hours like this.

Code:

Mem: 223984K used, 286588K free, 0K shrd, 28K buff, 203096K cached
CPU:   4% usr   8% sys   0% nic  86% idle   0% io   0% irq   0% sirq
Load average: 2.00 2.01 2.04 1/68 12077
  PID  PPID USER     STAT   VSZ %VSZ %CPU COMMAND
 8322  8321 root     S     112m  23%   4% ./cgminer -c /config/cgminer.conf
  326     1 root     S     3272   1%   0% /usr/bin/ntpd -p /var/run/ntp.pid -g
  341     1 root     S     3112   1%   0% /usr/sbin/lighttpd -f /etc/lighttpd.co
  334     1 avahi    S     2824   1%   0% avahi-daemon: running [knc-jupiter1.lo
 9390   323 root     S     2732   1%   0% /usr/sbin/dropbear -r /config/dropbear
  335   334 avahi    S     2728   1%   0% avahi-daemon: chroot helper
 8321     1 root     S     2616   1%   0% {screen} SCREEN ./cgminer -c /config/c
 9576  9575 root     R     2328   0%   0% top
 9575  9390 root     S     2328   0%   0% -sh
  323     1 root     S     2272   0%   0% /usr/sbin/dropbear -r /config/dropbear
 1775     1 root     S     2148   0%   0% {monitordcdc} /bin/sh /sbin/monitordcd
11709  1775 root     S     2016   0%   0% sleep 60
 1773     1 root     S     1908   0%   0% /sbin/getty 115200 ttyO0
 1774     1 root     S     1908   0%   0% /sbin/getty 38400 tty1
    1     0 root     S     1652   0%   0% init [5]
 1756     1 root     S     1496   0%   0% /usr/bin/monitor-pwbtn /usr/bin/factor

[markm]

I am using the executable ckolivas most recently posted about, supposedly 3.7.1

I don't have an ARM development machine nor know a thing about cross-compiling on Fedora 17 or 19 for ARM machines.

I have run over 48 hours before dying  sometimes, even over 72 maybe. Actually maybe it didn't die like this until they released a release that had a ckolivas version in it, hmm I wonder if that one is/was stripped? Hmm... I am not sure when I first encountered this "mystery dying" problem, seems liek it ran quite a few days originally.

I killed it to try running strip on it on the beaglebone, that is how I discovered they dont' seem to have strip on that. But now it shows in web page as 282 so maybe its not goign to lose much by using --lowmem option afterall. I do wonder how much RAM the debugging symbols are consuming though...

-MarkM-

[sickpig]

Yep it was out of memory, dmesg shows:

Code:

[292489.141503] Out of memory: Kill process 3622 (cgminer) score 188 or sacrifice child
[292489.149604] Killed process 3622 (cgminer) total-vm:189152kB, anon-rss:109744kB, file-rss:1652kB

-MarkM-

try to contact ckolivas and report what happened in the meantime switch back to a working version of cgminer.

I grabbed the ckolivas one because it was said to possibly fix cases where mining stops completely. which is what I had been seeing from the official release.

I hoped that this problem of mining being stopped completely according to the web page, which has turned out to be due to an out of memory condition, was the mining stopping completely problem that the wget grabbed version hoped to fix.

In other words, I didn't have a working version in the sense of one that did not mysteriously die after a day or two or few and hoped the wget would get me one that would not thusly die.

I am trying the latest wget with --lowmem now hopefully that won't die...

it would lower the amount of used mem since it tries to "Minimise caching of shares for low memory applications", it's worth a try.

But it seems to be settling at 278 GHashes whereas I had been getting usually 283, so maybe putting another stick of RAM or larger stick of RAM in the damn beaglebone might be better if --lowmem loses hashing power. I wonder if the damn beaglebones can handle more RAM? Maybe KnC skimped on RAM too quickly without considering that software tends to get bigger all the time...

not as easy as it seems, let say it's not like adding a module of RAM to a desktop
see https://groups.google.com/d/msg/beagleboard/aWpARi3fdCo/GWMpkFIOH-QJ for more info.

Or maybe I should kill the stupid webserver, if I can ssh what for do I need a webserver on the thing anyway? Hmm...

Also, I looked at the file I had used wget to get with "file" command on an X86_64 , and that claims its an ARM executable that is not stripped. Wouldn't it maybe use less RAM if it were stripped? Does it make much sense to use a not-stripped executable on an "imbedded device" type thing like this where RAM is at a premium and no debugger is installed? Any chance of a stripped version, ckolivas? (My X86_64 strip doesn't even recognise the thing as an executable and the miner does not have strip installed...)

stripping cgminer executable shouldn't  do any harm. the fast way u can strip it without waiting for ckolivas is install strip cpmmand into your mienr a do it directly from an ssh session.

-MarkM-

[/quote]

having said that I've noticed a slightly increase in system avg load in the latest version of the firmware. on my miner with 0.95 sys load was ~1.95-1.98 with 0.98/0.98.1 I get something like 2-2.1 .

while using 0.95 I've tested bfgminer and the sys load lower to 1.55 unfortunately I don't have time to repeat the test with 0.98.1, but if you feel brave you could it for me  using this recipe http://codepad.org/QKSeO5zh

[markm]

The miner has installers and such? What do you mean by install strip command?  yum install strip or apt-get install strip or apt install strip type of thing?

I figured off the cuff, wild assed guess, that if strip isn't there likely things like package managers / installers likely aren't either...

What is that recipe for? Cross-compiling on some other machine? Or you run that on your miner?

-MarkM-

[sickpig]

The miner has installers and such? What do you mean by install strip command?  yum install strip or apt-get install strip or apt install strip type of thing?

I figured off the cuff, wild assed guess, that if strip isn't there likely things like package managers / installers likely aren't either...

What is that recipe for? Cross-compiling on some other machine? Or you run that on your miner?

-MarkM-

the recipe contains instructions to compile bfgminer directly on the miner.

to install strip just execute the first 6 line of the recipe and then just issue:

opkg install binutils

(or whatever is the package that contain the strip binary in an angstrom linux distro)