Topic: 1 TH/s-miner with only 16 GH/s (Read 3483 times)

Ohh by the way
What is the temperature on that board (nr 4)?
Check the other values and see so there isnt any strange things that can lead you in any direction...

In 10 hours i am in front of my machine again and can compare numbers...

// U

Okidoki

I had some problems the first start and when i changed IP-adress.
Sometimes Asic 4 is not showing up for me, but perfoming a coldstart normally fix it.
You can also check the lights on each asic blade, there is around 10 green leds in a row and they should be green. If some of them are red that means that some chips are offline because of a failure. It could happen when power is to low or something else spoky :-)
Just restart the machine and hopefully all go green.
I have this problem on another miner, sometimes i need 3-4 restart until its all green.

If the problem persist it could be a fault in the machine and you need to get it fixed.

// U

Hi Urbie.  Thanks for the advise. I tried what you said, but alas still having troubles with my machine.

You can see in the pic, error rate on module #4 is very high, and therefor the speed is only 700-800 GH/s.  Speed on module #4 is obviously wrong.
https://i.imgur.com/s6fcfrv.jpg

Also I have a customer with another machine where the module #4 is not showing up at all, so he's only getting 650-750 GH/s.
https://i.imgur.com/ANd0gue.jpg

Any thoughts?

Hi,

I fixed this error on my machine.
I have found out that you should never use 1024 on the machine side.
Normally CGMINER and pool communicates about best difficulty,  and to start at highest level is a very bad idea from machine side.
Set the machine to 256 (default) and the it works flawlessly.

If you wanna raise difficulty set the minimum diff in the pool. The parameters is called "minimum difficulty" and the difficulty starts at the value you choose. If yu choose 16 it will very soon automaticly raise.
In my 1 Th machines i use 1024 "minimum diff" and in my 200 Gh machines i use 512 or 256.
It doesnt effect the hashpower that much but the 1 Th machines works more stable with 1024 and it also minimize network traffic.

So problem solved :-)

// Urbie

Hello.  Any luck with this?  One of these machines I ordered is showing only 600-700 GH/s on BTCGuild dashboard.  The machine itself thinks its doing 936 GH/s.  I have the minimum difficulty on BTCGuild and the machine set to 1024.  Thanks for any help or pointers.

getwork has nothing to do with modern mining. It's almost certainly just a minimum share diff in their broken driver implementation, as discussed before.

I have this machine and it is getting 1.05 terrhash per second no problem. I used btc guild at difficulty 1024. No problems.

My guess is that lower difficulty settings result in so many hits that it cannot communicate them fast enough. At 1024 I am getting a hit every few seconds. It may also be optimized for that difficulty.

Thanks to the guy who told me the ssh username and password!

I think its honestly just a compatibility issue with the getwork software poolside.

The miner is showing actual hashrate clientside and then share hashrate poolside and its submitting shares at a rate of 16gh aka 1.06% share submission rate(99% stales/rejects). My guess...

That's sound like a true ! I guess that company would like to get money despite everything

I think the saler cheat you

You should return it

Nope. I've ordered with Bitmine (yeah, beat me), so I would recognize them :-)

Could you have gotten one of these: http://bitmine.ch/?p=5178

Yeah it may have internal firmware diff mining and if it is mining at 1k diff (i.e. only returning nonces at 1k diff) while the pool is asking for 10diff, it will effectively be returning only 1/100 of the nonces the pool wants and this would make it appear 100 times slower than it is.

That is weird.
Have you tried a packet sniffer to see if it is mining somewhere else as well?
I have heard complaints of some of the miners that are run with Pi's using a small amount of mining power to mine to a different pool with out showing up.
I wonder if that is going on here but it malfunctioned and instead is mining mostly to the hidden setup.

why wouldnt the company tell you about this before figuring it your self tho?

It may be a problem with the machines ability to handle low difficulties, causing it to basically create a backlog it can't escape from.  Have you tried setting your difficulty to 1024 on another pool?  You're going to be running at 512-1024 diff no matter what, standard variable difficulty on all pools is going to put you in that range if your machine is functioning properly.

tststststs... as an old bash-er, I do not like GUIs :-)

Or Wireshark for people who like GUI.

WU 12/m means 829MH/s ... so not what it should be.
for 1TH/s WU should be around 14000/m

A1 chips need a high diff or a pool that supports 0 diff (auto).

I run my Hex8A1 (8x A1 chips) on BitMinter and BTCGuild.

I've also thought the new Dragon miner (4 boards x 8 chips) may be running too low voltage which would result in high hardware errors/rejects.

Well it shouldn't be hard to see where the device is actually mining.
A tcpdump on the network should show where packets are going to and coming from.
More so, if you start the log when it is first powered on, it should show all pools and accounts it is connecting to.

Interesting for me is the fact that there is a lot of references to bitmine in the history. For example here in the history of user "pi":

  173  gvim bitmine.h
  392  cd cgminer-bitmine-A1-scratchpad/
  395  cd cgminer-bitmine-A1-scratchpad/
  398  cd cgminer-bitmine-A1-scratchpad/
  399  gvim driver-SPI-bitmine-A1.c
  402  cd lichen/cgminer-bitmine-A1-scratchpad/
  404  cd lichen/cgminer-bitmine-A1-scratchpad/
  406  cd lichen/cgminer-bitmine-A1-scratchpad/
  408  cd lichen/cgminer-bitmine-A1-scratchpad/
  445  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  447  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  454  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  457  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  460  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  464  gvim ./lichen/old/cgminer-bitmine-A1-scratchpad/driver-SPI-bitmine-A1.c
  466  gvim ./lichen/old/cgminer-bitmine-A1-scratchpad/run.sh
  469  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  474  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  480  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  487  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  488  gvim driver-SPI-bitmine-A1.c
  495  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  509  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  519  gvim /home/pi/lichen/old/cgminer-bitmine-A1-scratchpad/run.sh
  520  /home/pi/lichen/old/cgminer-bitmine-A1-scratchpad/run.sh &
  521  sudo /home/pi/lichen/old/cgminer-bitmine-A1-scratchpad/run.sh &
  523  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  548  cd /home/pi/lichen/old/cgminer-bitmine-A1-scratchpad
  550  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  562  cd lichen/old/cgminer-bitmine-A1-scratchpad/
  637  cd lichen/old/cgminer-bitmine-A1-scratchpad/


Now given the fact that Bitmine in Switzerland can't deliver, but here obviously some software from Bitmine pops up in a Chinese machine which is sold from a Swiss company, I guess I'll use some forensic tools to examine this issue further...

I'm running one of my micro-pool (hey, I like this term!) with ten Antminer on Eligius with no problem. However, Eligius, ghash and others don't work with this machine, only BTCguild - at the moment...

Don't know if we have the same machine, but here the default user is pi, password is raspberry. sudo su makes you root where you can set your root-password. At least worth a try...

Okay, BTCGuild with 1024 min diff works with somewhat over 1 TH/s. However, this means I can't put it on other pools - unless I want my heater to run for 16 GH/s...

You are right that is bullshit. There might be 0.1 % difference of different pools. But That much different is bullshit.

Hmm, my miner.php there is showing you are mining on Eligius ... which has been DDOSed for a few? days.

Although they like to say their magical unicorns can solve all DDOS problems, maybe that's the reason?

At a guess, I'd say that's not the case. I can't think of any reason it would only hash at BTC Guild at a high min diff.

A device that can only hash in a specific pool is definitely not worth $5000.

Did you try BTCGuild with 1024 min diff? You'd need to run it for a day, but at least you'd know if they were completely full of shit or just a little full of shit.

I have the exakt same issue.
I received my machine 1 day ago and after setting up i was happy. Until today.
I normally use Slush pool. I now tried Elegius and GHashIO with same result.
Besides that,  I can´t login with SSH, i dont have the correct user and password... Anyone know the default password? No info in the box... The webinterface  doesn´t require user and password.
In the display and also in webinterface all looks perfect. 1.02 Th and WU 12m somethings. Looks OK.
BUT. No money, no hashrate on pool side. Only between 1-16 Gh. Mostly 1 Gh.

I am totally out in the blue here.
Any suggestions?

Thanks

Urbie

Since cgminer is released under a free license that insists all code modifications distributed in binary form must be available on request, since you have received a cgminer binary that is modified from the original, you can ask them for the source code to their modifications and they will be obliged to provide their source modifications within 30 days or they will be in violation of the license agreement and not be allowed to further distribute their binaries or be up for legal action.

See:
https://www.gnu.org/copyleft/gpl.html

That sounds like a bogus claim, cgminer doesn't care what pool it is, it passes the work to the driver which sends it to the ASIC, mine for another 30 minutes to an hour and if the speed is still 16gh/s, I smell a scam.

Yup, and they claim that I only can use the machine with BTCGuild with the setting 1024/1THs as minimal diff. So this would be a limitation I can't accept.

Now that's becoming interesting. When I look at the history in the bash, I see that the name "Bitmine" appears again and again. However, the machine itself is the one sold by the "LK Group" in China (http://www.lketc.com). But the interface is produced by another company in Switzerland (roxas.ch).

So I'll dig a bit deeper into this during the weekend. What would be interesting is if anybody who bought from LK in China has the same issue...

have you tried contacting the company about your findings?

Naugh, currently more 1 TH/year... :-/

When I ssh into the miner and try to start cgminer, I just get a segmentation fault if I'm root. When I'm just a normal user, it sometimes starts when I put in sudo cgminer.... however, the output is not verbose, and if I try to start cgminer with the --verbose option, it doesn't start at all. However, it starts perfectly when I use the web-interface. ps -Af shows the following:

 sudo ./cgminer -o stratum+tcp://stratum.mining.eligius.st:3334 -u xxxxxxxxxxxxxx -p 0 -o stratum+tcp://uk1.ghash.io:3333 -u xxxxx.6 -p 0 -o stratum+tcp://uk1.ghash.io:3333 -u xxxxx.8 -p 0 --A1Pll1 1000 --A1Pll2 1000 --A1Pll3 1000 --A1Pll4 1000 --A1Pll5 1000 --diff 16 --api-listen --cs 8 --stmcu 0 --hwreset --api-allow W:127.0.0.1

It seems as if they have produced their own little hacks for running cgminer....

But it is not 1TH/s is it? 

Only thing I can suggest is to SSH into the miner and see what cgminer is actually showing rather than the web front end. Because from the stats on the web it looks good.

For illustrating the situation here's a screenshot from the miner's gui:




However, after 20 minutes, Eligius shows me the following:



Now, after 30 minutes, I'm down on 6 GH/s.

Is there anything you guys who have more experience see?

Thanks

Tried Eligius.st and got there about 16 GH/s. At btcguild only 12, ghash.io about 14 and some smaller pools showed similar results. If it were only one pool showing such a low figure, I'd say it is a question of the miner not being compatible with the pool. But the fact that all pools show such low figures lets me guess that the miner has some issues..

Which pools have you tried? Try eligius.st

About 30 minutes on the first two pools, about 15 minutes on the other.

How long have you left it running on the pool before ascertaining that it only hashes at 16GH?

Hi all,

I just received a 1TH-miner. It works well. At least, that's what it looks like. On the interface, it pretends to work with more than 1 TH/s. However, on all pools i connected the machine I never got more than 16 GH/s. Has anybody an idea where I should start to investigate what's wrong with this machine?

It basically is this one: http://coinsfortech.com/shop/coincraft-1ths-28nm-asic-bitcoin-miner-with-power-supply/

Thanks for every hint!