Miner- Zogminer: Linux SILENTARMY

[scavern]

Bugger, anyone else still getting seg faults?

[defekt]

Awesome thanks! -S for each card got me going.

[scavern]

Looks like Omar is busy with build #103 Travis CI - Test and Deploy Your Code with Confidence

[scavern]

I got seg faults after about 5 minutes of mining on 1 x r9 380. so back to eth for me as well

[zcash111]

I got seg faults too.

[neo]

Who’s account is this? Im mining to it.

./zcash-miner -G -stratum=“stratum+tcp://us1-zcash.flypool.org:3333” -user=t1PGgRgVQ14utsD7mp2dzGdykTDFUCKzPQ5.rig1 -printtoconsole -S=1

[trolloniex]

voxelot:

ZEC: t1PGgRgVQ14utsD7mp2dzGdykTDFUCKzPQ5

BTC Donations: 1A67XmfyukPNbrMmckBToE8xZPesuM6Ti1

ETH: 0x82a642cAFD6bFc228C7B5Ea0394A0Ad702A7c03D

[neo]

Good. its the correct account. Thanks

[farallon]

@trolloniex

screen -a ctrl+a+d and run different instances ?

Thanks much for all this work, just sent some BTC to you and @omaralv. Does @Age have BTC address? If he does please place it on your first post (post #1) to make it easy to find along with the other addresses.

After much frustration, I finally got my rig working on Stratum. Here is some info that might be interesting and/or helpful to others:

• For pool mining, I compiled this code: GitHub - nginnever/zogminer: an OpenCL Zcash GPU miner
• This post was very helpful to get it running: Miner- Zogminer: Linux SILENTARMY - #516 by Age
• My rig: 6 R9 390 cards (4 4gb, 2 8gb), Ubuntu 14.04
• pool: supernova
• I’m running 2 processes per card on my 4GB and 3 process on my 8GB. For each process I just open up a dedicated terminal window and run the miner targeting a particular display: (is there a better way to do this?)

~/zogminer/src/zcash-miner -G -S=3 -stratum="stratum+tcp://zec-eu.suprnova.cc:2142" -user=user1 -password=password1"

• They all get about 38 - 39 H/s. I have not tried optimizing yet. I’m just happy they are running!

• I’m using display 0 also as my console, This is the only one that has crashed so far. Perhaps it is because I have only 4GB on this card?

• I’m using 2 power supplies: drawing 500 watts for my mobo + 2 cards and 890 for the other 4 cards.
  My ~/.zcash/zcash.conf is:

  addnode=mainnet.z.cash
  rpcuser=username
  rpcpassword=password
  genproclimit=2
  gen=1
  showmetrics=1

I’ve been running for about 12 hours now and it has been stable, except for Display0, as mentioned above.

For those getting very low H/s (like under 1H/s) it is likely you are only mining on your CPU and are not yet set-up correctly. At the time I was working on this, this project works for pool mining on mainnet and the other project that supported solo mining (GitHub - nginnever/zcash: zcash) only worked only the betanet. I was very confused by the two projects, anyway this might have changed by now…

All this has been very educational for me. What a ride!

Good work guys!

[osnwt]

USB flash drive boot, Ubuntu 14.04 LTS, fglrx, headless.

• 390X overclocked, 2 threads/card, 22+22=44-45 Sol/s, 4 cards, 12 hours, no issues.
• 390+390X overclocked, 2 threads/card, 44-45 Sol/s. 2 cards total, 12 hours, no issues. This rig also has 3 older cards (7950x2, 7970x1) that run eXtremal miner. I’ve heard zogminer does not work on old cards, but did not try yet. Hope to try very soon after payment from the pool.

Same Linux 14.04, amdgpu-pro.

• 470X 4GB (Elpida), 2 threads/card, 36 Sol/s, single card, 12 hours, no issues (headless).
• 470X 4GB (Hynix), 6 cards: no go (segfault on start).
  SAME USB DRIVE as above. Same miner binary as with fglrx above. Needs investigating since clinfo also crashes. Not a miner fault. I have issues with Claymore ETH miner as well on this system if run more than single GPU. Probably my fault somewhere, will do clean install. Windows runs Claymore miner on this with no issues (not a hardware fault).

I wrote simple bash script that runs many threads in background on the same console with given -S=n args, watches all processes and restarts missing in case of single thread failure (have not seen yet, zogminer seems to be very stable for me). Ctrl-C kills all processes as well before exit. Every few seconds in also logs running process PIDs. It runs zogminer for 12+ hours with no issues.

[osnwt]

Anybody runs zogminer on Windows? Should I try to build?

Still there are issues with undervolting new RX cards with amdgpu-pro drivers on Linux (no tools yet, BIOS changes do not undervolt memory controller for some reason). So maybe run on Windows until some Linux tools created.

[cascadiacrypto]

@voxelot

Zcash GPU Open-miner captains log 1.0.0:

As you can probably see by now, we are pleased to announce that we are steadily mining with multiple silentarmy threads over 6 GPU’s both solo and pool. We think we have everything cleaned up and working well enough to be used. A few more improvements can be expected over the next few days and a PR made to zcash for the solo mining section. The zogminer repo will remain a pool miner to use and could see big improvements over time.

At this point I feel about finished. This project has completed its goals and even exceeded them greatly. We were able to get open source software hashing faster and more stable than any proprietary software being sold. Cheers!

Congratulations to you, Omar, Age and everyone else involved! Your lack of sleep and round the clock dedicat to the open source miner effort as well as your ability to co-ordinate a global/multi timezone based team, was a thing to behold!

Out of your effort the beginnings of a mutually helpful ZCash mining community is emerging.

Of all the entries in the Open Source Miner Competition yours is most complete and should in my honest opinion, easily win the GPU miner reward.

But I’m not a judge. If I were, I’d vote for you guys though.

This is an exciting time and I can verify your miner works better than any thing else out there including closed-source/proprietary ones.

Surpassing Jtoomim’s inhouse secret miner through rigorous coding and testing is a profound achievement and victory for open source and a secure decentralized network in its infancy.

Congratulations! Now get some well deserved rest

[guizmo]

I get an errer trying to build it. Weird since i already have another rig working with no problems… anyone can help?

/usr/bin/ld: cannot find -lOpenCL
collect2: error: ld returned 1 exit status
Makefile:3188: recipe for target ‘zcashd’ failed
make[2]: *** [zcashd] Error 1
/usr/bin/ld: cannot find -lOpenCL
collect2: error: ld returned 1 exit status
Makefile:3162: recipe for target ‘zcash-gtest’ failed
make[2]: *** [zcash-gtest] Error 1
/usr/bin/ld: cannot find -lOpenCL
collect2: error: ld returned 1 exit status
Makefile:3103: recipe for target ‘test/test_bitcoin’ failed
make[2]: *** [test/test_bitcoin] Error 1
make[2]: Leaving directory ‘/home/athos/Desktop/zogminer/src’
Makefile:7546: recipe for target ‘all-recursive’ failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory ‘/home/athos/Desktop/zogminer/src’
Makefile:645: recipe for target ‘all-recursive’ failed
make: *** [all-recursive] Error 1

[osnwt]

Make symbolic link to AMD’s libOpenCL.so and libOpenCL.so.1 to /usr/lib/ - so linker will be able to find this lib. There are better solutions but this one works and easy to use.

[jack008]

is this miner good for AMD cards only? I have an old NVIDIA Quadro card

[voxelot]

@jack008

is this miner good for AMD cards only? I have an old NVIDIA Quadro card

Unfortunately silentarmy kernel is built to work well with AMD hardware and doesn’t run well or at all on nvidia. BUT, I just read mrb is trying to get his hands on some nvidia hardware to fix this.

[guizmo]

Thanks man. As soon as I read your post got everything solved.

[hme]

Im guessing this does not work in Ubuntu 16.10, as I cannot find any way of installing amdgpu-pro in 16.10…
Is it possible to use the open source drivers?
The program compiles but gets stuck:
The card is a stock rx480 - 4gb
“”"
Zcash Miner version v1.0.0-beta2-ad6750f (2016-10-31 10:17:41 +0100)
Running the test miner
[OPENCL]:Found suitable OpenCL device [AMD POLARIS10 (DRM 3.3.0 / 4.8.0-26-generic, LLVM 3.8.1)] with 1073741824 bytes of GPU memory
[OPENCL]:Using platform: Clover
[OPENCL]:Using device: AMD POLARIS10 (DRM 3.3.0 / 4.8.0-26-generic, LLVM 3.8.1)(OpenCL 1.1 Mesa 12.0.3)
[OPENCL]:input.cl:739:5: warning: implicit declaration of function ‘mem_fence’ is invalid in C99
AMDGPUMCInstLower::lower - Pseudo instruction doesn’t have a target-specific version: 146
Running Equihash solver with nNonce = 0000000000000000000000000000000000000000000000000000000000000000
Solver took 0.00 Mcycles
Running Equihash solver with nNonce = 0000000000000000000000000000000000000000000000000000000000000001
Solver took 0.00 Mcycles
Running Equihash solver with nNonce = 0000000000000000000000000000000000000000000000000000000000000002
Solver took 0.00 Mcycles
“”"
This last block repeats counting upwards indefinitely, but the solver does not seem to do anything
Card does not get hot ect…

Anyone got a simmilar setup running?

[voxelot]

@hme

[OPENCL]:Found suitable OpenCL device [AMD POLARIS10 (DRM 3.3.0 / 4.8.0-26-generic, LLVM 3.8.1)] with 1073741824 bytes of GPU memory

Hmm i develop on ubuntu 16.04 with amdgpu-pro and an rx480 so our setup is the same. One thing I notice is that opencl thinks that your card only has 1G of memory on it. That should be a problem.

[Louie2001912]

@neo

Who’s account is this? Im mining to it.

./zcash-miner -G -stratum=“stratum+tcp://us1-zcash.flypool.org:3333” -user=t1PGgRgVQ14utsD7mp2dzGdykTDFUCKzPQ5.rig1 -printtoconsole -S=1

LOL did someone get FUCK in their address. That’s awesome.