Pages:
Author

Topic: mkxminer: Fast Lyra2Z/Lyra2rev2 miner for AMD GPU (ZCoin, GINcoin, VERT, MONA... - page 11. (Read 52724 times)

jr. member
Activity: 74
Merit: 1
Anyone managed to pull off more than 55 Mh/s with Vega?

This miner considered to be the most profitable for non-ETH only GPUs.. but Fury X giving the same hashrate as brand new Vega? Sounds uncool

Think about it for a moment...the only major differences are an increase in memory from 4 gb to 8 gb and a few architectural tweaks...its more evolution than revolution in design. vega 2 at 7nm is where it will get interesting...
newbie
Activity: 53
Merit: 0
Anyone managed to pull off more than 55 Mh/s with Vega?

This miner considered to be the most profitable for non-ETH only GPUs.. but Fury X giving the same hashrate as brand new Vega? Sounds uncool
jr. member
Activity: 74
Merit: 1
trying to get mkxminer working on suprnova verge and getting an error stating error 500 and constant connecting and reconnecting...I'm using windows 10 64 bit and have 3 xfx 570's and one xfx 580 black. any idea what the hell wrong i'm doing wrong in my batch file or is there a problem with this miner xvg on suprnova. I see that they don't have any example miner config for amd cards for lyra2v2.
newbie
Activity: 4
Merit: 0
Ubuntu x64 mining from scratch step by step guide

Code:
sudo dpkg --add-architecture i386
sudo apt-get update
sudo apt-get install libc6:i386 libstdc++6:i386

Did this, but libOpencl.so.1 still not found. Reinstalled amdgpu-pro via its install script, but still not found.

Code:
sudo apt install libopencl1-amdgpu-pro:i386 opencl-amdgpu-pro-icd:i386 clinfo-amdgpu-pro:i386

This got it to run. However, it segfaults, but a forked process keeps running, doing nothing. Also, you need to implement something to disable extra nonce for p2pool:
Code:
$ ./mkxminer -o stratum+tcp://45.63.77.16:9181 -u /8+0.012 -p x -I 20
[09:56:20 - 0 days 00:00:00] Starting mkxminer 3.1.0...
[09:56:20 - 0 days 00:00:00] Non-standard pool: please consider using a recommended pool:
[09:56:20 - 0 days 00:00:00]   *      stratum+tls://stak.suprnova.cc:7710
[09:56:20 - 0 days 00:00:00]   *      stratum+tls://mona.suprnova.cc:3001
[09:56:21 - 0 days 00:00:00]   *      stratum+tls://vtc.suprnova.cc:5676
[09:56:21 - 0 days 00:00:00] More information: https://bitcointalk.org/index.php?topic=2360168
[09:56:21 - 0 days 00:00:00] GPU_MAX_ALLOC_PERCENT or GPU_USE_SYNC_OBJECTS environment variable is not set, performance may be worse!
[09:56:21 - 0 days 00:00:00] Connecting to stratum+tcp://45.63.77.16:9181 as /8+0.012...
[09:56:21 - 0 days 00:00:00] Difficulty set to 138.736
[09:56:22 - 0 days 00:00:01] Connected!
[09:56:23 - 0 days 00:00:02] --- Detected Platforms ---
[09:56:23 - 0 days 00:00:02] *0 AMD Accelerated Parallel Processing OpenCL 2.1 AMD-APP (2527.3)
[09:56:23 - 0 days 00:00:02] --- Detected GPUs ---
[09:56:23 - 0 days 00:00:02] GPU#0: Pitcairn
[09:56:23 - 0 days 00:00:02]
[09:56:23 - 0 days 00:00:02] Press 'q' to exit application
[09:56:23 - 0 days 00:00:02]
[1]  + 10987 segmentation fault  ./mkxminer -o stratum+tcp://45.63.77.16:9181 -u  -p x -I 20
$ echo $?
139
$                                                                                                              
[09:56:23 - 0 days 00:00:02] Pool stratum auth error: {
   "message": "Service not found"
[09:56:23 - 0 days 00:00:02] Disconnecting from stratum
[09:56:23 - 0 days 00:00:02] Difficulty set to 138.736
[09:56:25 - 0 days 00:00:05] Pool stratum auth error: {
   "message": "Service not found"
[09:56:25 - 0 days 00:00:05] Disconnecting from stratum
[09:56:33 - 0 days 00:00:12] Idle. No work from pools.
[09:56:56 - 0 days 00:00:35] Difficulty set to 102.023
[09:57:27 - 0 days 00:01:06] Difficulty set to 185.665
[09:57:29 - 0 days 00:01:08] Pool stratum auth error: {
   "message": "Service not found"
[09:57:29 - 0 days 00:01:08] Disconnecting from stratum
[09:57:59 - 0 days 00:01:38] Difficulty set to 144.004
[09:58:01 - 0 days 00:01:40] Pool stratum auth error: {
   "message": "Service not found"
[09:58:01 - 0 days 00:01:40] Disconnecting from stratum
> 0.0MH/s | Temp(C):                | | Fan: | HW: 0 | Rej: 0.0%
$
$ jobs
[1]  + segmentation fault  ./mkxminer -o stratum+tcp://45.63.77.16:9181 -u  -p x -I 20
$ fg
 > 0.0MH/s | Temp(C):                | | Fan: | HW: 0 | Rej: 0.0%
[1]  + 10987 continued  ./mkxminer -o stratum+tcp://45.63.77.16:9181 -u  -p x -I 20
\^\^C^\^C[1]  + 10987 quit (core dumped)  ./mkxminer -o stratum+tcp://45.63.77.16:9181 -u  -p x -I 20

And I can't seem to get it to run under gdb, not even with no args. It just "exits normally" with no output... do you have some debugger detection going on? Would love to be of more help, but paranoia in the way? This is a single r9-270, by the way, with amdgpu-pro version 17.50-511655.

If I use stratum+tls://vtc.suprnova.cc:5676 it does not segfault, but stratum+tcp://vtc.suprnova.cc:5678 does.

Update: If I leave it running for a while, it still segfaults eventually
Code:
[19:46:34 - 0 days 05:44:51] Accepted diff 16 share 04baae71 GPU#0 in 161ms
[19:46:36 - 0 days 05:44:53] Accepted diff 16 share 0b3c303e GPU#0 in 158ms
[19:46:46 - 0 days 05:45:03] Accepted diff 16 share 02e98493 GPU#0 in 161ms
[19:47:11 - 0 days 05:45:28] Accepted diff 16 share 0e50c72f GPU#0 in 162ms
[19:47:40 - 0 days 05:45:57] Accepted diff 16 share 0f94a123 GPU#0 in 163ms
[19:47:54 - 0 days 05:46:10] Accepted diff 16 share 0bf5efc3 GPU#0 in 163ms
[19:48:18 - 0 days 05:46:35] Can't receive server reply - error 301
[19:48:18 - 0 days 05:46:35] Disconnecting from stratum
[19:48:18 - 0 days 05:46:35] Devpool: connection interrupted
[19:48:18 - 0 days 05:46:35] Devpool: restarting...
[1]  + 23811 segmentation fault  ./mkxminer -o stratum+tls://vtc.suprnova.cc:5676 -u -p -I 21 --asm
[19:48:45 - 0 days 05:47:02] Idle. No work from pools.
[19:58:00 - 0 days 05:56:17] Can't receive server reply - error 301
[19:58:00 - 0 days 05:56:17] Disconnecting from stratum
[19:58:00 - 0 days 05:56:17] Pool: connection interrupted
[19:58:00 - 0 days 05:56:17] Pool: restarting...
[19:58:01 - 0 days 05:56:18] Difficulty set to 32
[19:58:56 - 0 days 05:57:13] Can't receive server reply - error 301
[19:58:56 - 0 days 05:57:13] Disconnecting from stratum
[19:58:56 - 0 days 05:57:13] Devpool: connection interrupted
[19:58:56 - 0 days 05:57:13] Devpool: restarting...
[20:08:31 - 0 days 06:06:47] Can't receive server reply - error 301
[20:08:31 - 0 days 06:06:47] Disconnecting from stratum
[20:08:31 - 0 days 06:06:47] Pool: connection interrupted
[20:08:31 - 0 days 06:06:47] Pool: restarting...
[20:08:31 - 0 days 06:06:48] Difficulty set to 32
[20:09:25 - 0 days 06:07:42] Can't receive server reply - error 301
jr. member
Activity: 156
Merit: 1
Is it possible to use this miner as a third party with Nicehash miner Legacy (1.8.1.Cool ?

You can point the miner at nicehash directly if you wish. Just choose the pool/stratum for the algorithm that you want to mine here - https://www.nicehash.com/help/which-stratum-servers-are-available

You can check your statistics and profits etc by looking up your btc wallet address on nicehash.

I'm still a noob at mining, how to "point it" at nicehash please ? Thanks.

you point your miner to
stratum+tcp://lyra2rev2.EU.nicehash.com:3347 (if you are in EU, or USA or...)

see
https://www.nicehash.com/algorithm/lyra2rev2

and as a wallet, you use your nicehash bitcoin address

example:
mkxminer.exe --asm -o stratum+tcp://lyra2rev2.EU.nicehash.com:3347 -u yourbitcoinaddress.optionalminer -p x -I 23

and remove .optionalminer if you don't want it
jr. member
Activity: 156
Merit: 1
Is it possible to use this miner as a third party with Nicehash miner Legacy (1.8.1.Cool ?

You can point the miner at nicehash directly if you wish. Just choose the pool/stratum for the algorithm that you want to mine here - https://www.nicehash.com/help/which-stratum-servers-are-available

You can check your statistics and profits etc by looking up your btc wallet address on nicehash.

I'm still a noob at mining, how to "point it" at nicehash please ? Thanks.

you point your miner to
stratum+tcp://lyra2rev2.EU.nicehash.com:3347 (if you are in EU, or USA or...)

see
https://www.nicehash.com/algorithm/lyra2rev2

and as a wallet, you use your nicehash bitcoin address
newbie
Activity: 84
Merit: 0
Hello, great job for making this miner, I get on my vega 56 rig hashrates as advertised.
Have problem with --exitsick parameter. One of my cards becomes "sick" but miner is not exiting, even using "q" to manually close it does not work, it's just continue to work without that one "sick" gpu.
Since miner does not quit I cannot use script to reset cards and start over.
I have to close miner by closing its window.
full member
Activity: 1120
Merit: 131
Is it possible to use this miner as a third party with Nicehash miner Legacy (1.8.1.Cool ?

You can point the miner at nicehash directly if you wish. Just choose the pool/stratum for the algorithm that you want to mine here - https://www.nicehash.com/help/which-stratum-servers-are-available

You can check your statistics and profits etc by looking up your btc wallet address on nicehash.

I'm still a noob at mining, how to "point it" at nicehash please ? Thanks.
full member
Activity: 1167
Merit: 209
Is it possible to use this miner as a third party with Nicehash miner Legacy (1.8.1.Cool ?
If an API is added to mkxminer, this miner will be added to Nicehash Miner Legacy Fork Fix
sr. member
Activity: 1246
Merit: 274
Is it possible to use this miner as a third party with Nicehash miner Legacy (1.8.1.Cool ?

You can point the miner at nicehash directly if you wish. Just choose the pool/stratum for the algorithm that you want to mine here - https://www.nicehash.com/help/which-stratum-servers-are-available

You can check your statistics and profits etc by looking up your btc wallet address on nicehash.
full member
Activity: 1120
Merit: 131
Is it possible to use this miner as a third party with Nicehash miner Legacy (1.8.1.Cool ?
newbie
Activity: 3
Merit: 0
This miner makes my PC crash. I get black screen and after 5 sec. my PC reboots.
full member
Activity: 241
Merit: 100
Is there a trick to get a 280x to mine with this on windows? Anything config I try crashes, even the most basic ones. Maybe because i'm using the blockchain driver ?
copper member
Activity: 5
Merit: 0
The ubuntu server "wget https://filebin.net/xs1yhv0qlpobge74/mkxminer310.tar.gz" did not work for me.  I had to copy over with a thumb drive from another machine. Otherwise with an R9 290 I got 19+ Mh/s basic and 25 Mh/s with --asm.  Thanks
newbie
Activity: 86
Merit: 0
tried again and again with diffrent and GPu Pro, Ellesmere is not Supported, it looks like your code only works for Older AMD Cards, its not Driver related Issue in Linux

Can you please fix this issue.


Hi!

Yes, I'm afraid Ellesmere is not supported on Linux yet. I'll look into that.

R9 390, 380X, 280X definitely work - I tested myself on Ubuntu 16.04 x64.
newbie
Activity: 86
Merit: 0
Ubuntu x64 mining from scratch step by step guide

*** Prerequisites:
2 USB flash drives (or 1 USB + 1 SSD/HDD), AMD videocard

*** Install Ubuntu Server
Download Ubuntu 16.04 x64 Server ISO image (I used 16.04.3 LTS from https://www.ubuntu.com/download/server).
Copy the Ubuntu installer to a USB. I recommend using the "Universal USB Installer".
Reboot with installation USB and the destination USB inserted (or just the installation USB and SSD/HDD connected).
Select install option.
Do a standard install to USB/HDD/SSD. Make sure "OpenSSH server" option is turned ON (it is needed to connect remotely via terminal later).
Remove installer USB and reboot.

*** Get Ubuntu x64 to work with 32 bit applications
(You can now login from from another terminal over SSH. For example Putty for Windows or the built-in Terminal / Hyper / iTerm2 for Mac, whatever..., even "Prompt" for iOS will do)
Login to your Ubuntu server box (ssh yourlogin@yourubuntuaddress)
Execute these commands:
Code:
sudo dpkg --add-architecture i386
sudo apt-get update
sudo apt-get install libc6:i386 libstdc++6:i386
now reboot your computer


*** Set up AMD driver
Download AMD driver. Now this becomes interest. There are a few versions out there to choose from and probably there will be even more by the time you read this.
Code:
wget --referer=https://support.amd.com https://www2.ati.com/drivers/linux/beta/ubuntu/amdgpu-pro-17.40-483984.tar.xz
or
Code:
wget --referer=https://support.amd.com https://www2.ati.com/drivers/linux/beta/ubuntu/amdgpu-pro-17.40-492261.tar.xz

install xz utils if you don?t have yet (required to extract the driver from .tar.xz)
Code:
sudo apt-get install xz-utils


Code:
tar -xJvf amdgpu-pro-17.40-483984.tar.xz
cd amdgpu-pro-17.40-483984

./amdgpu-pro-install
sudo reboot

Add your user to video group:
Code:
sudo usermod -a -G video $LOGNAME 
sudo reboot


*** Download & run mkxminer
Download mkxminer (links in bitcointalk thread). For example:
Code:
wget https://filebin.net/xs1yhv0qlpobge74/mkxminer310.tar.gz

Extract the archive, for example:
Code:
tar -xf mkxminer310.tar.gz

Test mkxminer:
Code:
./example.sh

If it's ok and the app is running - close it by pressing 'q'.
Now edit example.sh and add your mining address / pool config.

*** Possible problems when running mkxminer:

If you execute ./mkxminer and it just closes without printing anything, then you don't have 32 bit support on your 64 bit Linux.

* Error: ./mkxminer: error while loading shared libraries: libOpenCL.so.1: cannot open shared object file: No such file or directory
Reason: You don't have 32bit OpenCL installed. If you added 32 bit support (like written above), you may need to reinstall AMD drivers.

* Error: Frontend phase failed compilation.
Error: Compilation from LLVMIR binary to IL text failed!
Reason: the AMD driver is not compatible with mkxminer. Try another driver. You can see AMD GPU-PRO version by executing: dpkg -l amdgpu-pro
AMD RX480 and RX580 currently have this problem in Linux, I?m not sure how to fix it yet.

* Error: clGetPlatformsIDs error -1001, verify that drivers are installed properly
Possible cause: Have you added your user to "video" group and rebooted after that?

Happy mining!
newbie
Activity: 36
Merit: 0
tried again and again with diffrent and GPu Pro, Ellesmere is not Supported, it looks like your code only works for Older AMD Cards, its not Driver related Issue in Linux

Can you please fix this issue.
newbie
Activity: 36
Merit: 0
Quote
rror: Frontend phase failed compilation.
Error: Compilation from LLVMIR binary to IL text failed!
Reason: the AMD driver is not compatible with mkxminer. Try another driver. You can see AMD GPU-PRO version by executing:
Code:
dpkg -l amdgpu-pro

For example, amdgpu-pro-17.40-492261.tar.xz and amdgpu-pro-17.40-483984.tar.xz show this problem on RX480, but work fine on R9 390.

I running rx480 rigs and rx 570 and rx580,
Tried 4 diffrent amg gpu pro, from 16.40 to the newest and erhe blockchain, it always gives the same error, it detects the gpus but shows ellesmere is not supported obly GCN.

Regards
newbie
Activity: 86
Merit: 0
Hello miners!

Please reply if you were able to get mkxminer running on Linux.
I can see many people are having problems, but has anybody been successful?

I installed a fresh Ubuntu 16.04 x64  desktop on a USB drive and got mkxminer running in 5 minutes.
Here are the required commands: https://bitcointalksearch.org/topic/m.24057659

PS: I tried recompiling for x64, but then 32bit OpenCL kernels don't work. So probably I'd need to make some changes to the ASM kernels to port them to 64 bit. Which is quite unfortunate, because everything works just fine on Windows 32/64 bit.

PPS: RX480/RX580 seem to be failing on Ubuntu x64 ("Compilation from LLVMIR binary to IL text failed"), not sure why but can reproduce myself. Will look into that.

Thanks!



Note: mkxminer is a 32-bit application. So if you are using a 64 bit application, like Ubuntu 16.04 x64, then you need
to install some some required libraries. Execute these commands in a terminal in case you're on Ubuntu x64:
Code:
sudo dpkg --add-architecture i386
sudo apt-get update
sudo apt-get install libc6:i386 libstdc++6:i386

If you execute ./mkxminer and it just closes without printing anything, then you don't have 32 bit support (see above).

Error: ./mkxminer: error while loading shared libraries: libOpenCL.so.1: cannot open shared object file: No such file or directory
Reason: You don't have 32bit OpenCL installed. If you added 32 bit support (like written above), you may need to reinstall AMD drivers.

Error: Frontend phase failed compilation.
Error: Compilation from LLVMIR binary to IL text failed!
Reason: the AMD driver is not compatible with mkxminer. Try another driver. You can see AMD GPU-PRO version by executing:
Code:
dpkg -l amdgpu-pro

For example, amdgpu-pro-17.40-492261.tar.xz and amdgpu-pro-17.40-483984.tar.xz show this problem on RX480, but work fine on R9 390.
newbie
Activity: 141
Merit: 0
Hi,

Can someone put me in right direction, why this might happen with miner.

Vega's 64 - 3 in Opteron 2218 (2.6Ghz) 8 RAM

Miner show following stat:

> Off 60.9MH/s 61.6MH/s 64.2MH/s | Temp(C): 56 40 43 47 | Fan: 39% 68% 0% 0% | HW: 0 0 0 0 | Rej: 0.0%

1 card - Off, this is how it should be this is old 7950, doesn't work good in this config.
3 Vega's show different speed, why this might happen?

Also, anyone experienced issue with Blockchain AMD drivers/Radeon Setting... Not even open to make any changes in WattMan.

Thanks.

I'm not sure if a 7950 is supported by this miner or not. I know that the 280X is supported, which is just a refreshed 7970, so it should be possible to have it work, but it might require some changes from the developer.... in any case, please share your .bat file so we can see how you have it setup.

Are all 3 of your Vega GPUs running at the exact same clocks?  If they're different models they may get different hashrates, and the good old silicone lottery will make even the exact same models hash a little differently in many cases.


Here is batch:

:restart
cd C:\Mining\mkxminer301\

devcon.exe disable "PCI\VEN_1002&DEV_687F"
devcon.exe disable "PCI\VEN_1002&DEV_679A"

timeout /t 10
devcon.exe enable "PCI\VEN_1002&DEV_687F"
devcon.exe enable "PCI\VEN_1002&DEV_679A"
timeout /t 5

setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100

OverdriveNTool.exe -p2vega_64 -p3vega_64 -p4vega_64

timeout /t 5

mkxminer.exe --exitsick --asm -o stratum+tcp://lyra2rev2.usa.nicehash.com:3347 -u x -p x -I 23 -d 1-3

goto restart

2 Vega's are same AMD (XFX) with Liquid cooling, 1 Vega is ASUS Air... running on same config via OverdriveNTTool (1409;1100)

Name=vega_64
GPU_P0=852;800;0
GPU_P1=991;900;0
GPU_P2=1084;950;0
GPU_P3=1138;1000;0
GPU_P4=1200;1050;0
GPU_P5=1401;1100;0
GPU_P6=1536;1150;0
GPU_P7=1409;950
Mem_P0=167;800;0
Mem_P1=500;800;0
Mem_P2=800;950;0
Mem_P3=1100;915
Fan_Min=4900
Fan_Max=4900
Fan_Target=75
Fan_Acoustic=2400
Power_Temp=85
Power_Target=0

Sapphire 7950 is actually working, but became sick very fast together with Vega's, also mining is not stable... rig might hangs in any time. So I push 7950 to equihash on same rig and everything seems to be stable, except :-) hashrate and radeon settings utility is dead...

Thanks.
Pages:
Jump to: