Hi,
About Error -1 when calling clGetDeviceIDs for number of devices issue :
Click Show hidden devices on the View menu in Device Manager and remove the unused gpu drivers (intel HD driver in my case), Reboot.
I did this on 2 rigs and was OK
Topic: Wolf's XMR GPU Miner - page 6. (Read 117592 times)
Wolfie 
Do you also have a Windows CPU XMR miner?
Do you also have a Windows CPU XMR miner?
How do i use it in linux with suprnova?
How do i use it in linux with dwarfpool?
How do i use it in linux with dwarfpool?
wolf -- can i put a failover pool in the same command line? using your latest v0.4/Linux/SimpleMinerOS. Thanks
No clue how to run this miner 
start.bat not working, config not working, miner.exe not working, I'm trying to mine Bip coin
start.bat not working, config not working, miner.exe not working, I'm trying to mine Bip coin
@Wolf0 kudos on the nice miner!
https://bitcointalksearch.org/topic/m.17548550
can you maybe add option to select OpenCL platform?
I'd like to run it on my 6*rx4 (including 2*480), but it's anniversary with integrated intel enabled (and genoil does have such switch, that I was using in early ZEC) otherwise there's no cards.
also does anyone know good params for 370?
I'm getting 470+ with:
"threads": 2,
"rawintensity": 380,
"worksize": 10
https://bitcointalksearch.org/topic/m.17548550
can you maybe add option to select OpenCL platform?
I'd like to run it on my 6*rx4 (including 2*480), but it's anniversary with integrated intel enabled (and genoil does have such switch, that I was using in early ZEC) otherwise there's no cards.
also does anyone know good params for 370?
I'm getting 470+ with:
"threads": 2,
"rawintensity": 380,
"worksize": 10
Hi,
I have 3 rigs, 2 x 6xRX480-8gb (H97 anniversary, AMD driver 16.9.2 and 16.11.4) and 1 x 2xRX470 (B85M-G43, AMD driver 16.11.4).
With the 2xRX470 rig, the Wolf's miner works fine.
1) with claymore's I have 1380h/s, with wolf's 1425h/s ( Thanks Za1n for config)
2) Power on the wall 330w --> 290w
Thanks Wolf.
All rigs use Windows 8.1 Embedded Industry Pro, same AMD drivers (tested : 16.9.2 and 16.11.4) but ... same issue with my 2 x 6xRX480-8gb rigs : Error -1 when calling clGetDeviceIDs for number of devices.
I used 6xgpu_mod for 2 x 6xRX480-8gb rigs ( https://lbr.id.lv/6xgpu_mod/6xGPU_mod_download.html)
I have 3 rigs, 2 x 6xRX480-8gb (H97 anniversary, AMD driver 16.9.2 and 16.11.4) and 1 x 2xRX470 (B85M-G43, AMD driver 16.11.4).
With the 2xRX470 rig, the Wolf's miner works fine.
1) with claymore's I have 1380h/s, with wolf's 1425h/s ( Thanks Za1n for config)
2) Power on the wall 330w --> 290w
Thanks Wolf.
All rigs use Windows 8.1 Embedded Industry Pro, same AMD drivers (tested : 16.9.2 and 16.11.4) but ... same issue with my 2 x 6xRX480-8gb rigs : Error -1 when calling clGetDeviceIDs for number of devices.
I used 6xgpu_mod for 2 x 6xRX480-8gb rigs ( https://lbr.id.lv/6xgpu_mod/6xGPU_mod_download.html)
It very likely has to do with the openCL platform, as the integrated Intel gpu could still be present. is there an option to select another openCL platform?
Oh this miner is so nice.
increase in speed and decrease in power consumption when switching from claymore.
Thanks Wolf0!
Only thing is i get "lowdifficultyshare" on nanopool, i think this might be due to some underclock of core and/or overclock of memory?
threads 2
rawintensity 504
workload 4
running 5x 470 4gb nitros
increase in speed and decrease in power consumption when switching from claymore.
Thanks Wolf0!
Only thing is i get "lowdifficultyshare" on nanopool, i think this might be due to some underclock of core and/or overclock of memory?
threads 2
rawintensity 504
workload 4
running 5x 470 4gb nitros
thanks for sharing.
Oh, my HD7950 just about burst into flames! Minimum fan speed is 60% for that old dog. I'm guessing the number of transactions per second has skyrocketed, are we under attack or just really popular?
edit:
I can smell piney resin enamel from the inductor windings vaporizing combined with your classic electrical fire odor. Yeah that puppy just got hot...
edit2:
The RX480 seems to be drawing considerably more than the manufacturers claimed 110W, I can now smell a hint of pine coming from my AX1200 PSU.... The GPU is however chilling.
edit3:
Insane, I had to shut the HD7950 down, the miner's AMD dual core CPU and PSU started bellowing toxic fumes. Both too hot to touch. I feel queezy.
WTF LOL
I'll try again later after it cools down and smell dissipates with CPU mining DISABLED!!! WOW I CAN"T BELIEVE THAT!!!
edit4:
Mining XMR is NOT for the faint of heart. Nobody is going to crack cryptonight, ever.
edit:
I can smell piney resin enamel from the inductor windings vaporizing combined with your classic electrical fire odor. Yeah that puppy just got hot...
edit2:
The RX480 seems to be drawing considerably more than the manufacturers claimed 110W, I can now smell a hint of pine coming from my AX1200 PSU.... The GPU is however chilling.
edit3:
Insane, I had to shut the HD7950 down, the miner's AMD dual core CPU and PSU started bellowing toxic fumes. Both too hot to touch. I feel queezy.
WTF LOL
I'll try again later after it cools down and smell dissipates with CPU mining DISABLED!!! WOW I CAN"T BELIEVE THAT!!!
edit4:
Mining XMR is NOT for the faint of heart. Nobody is going to crack cryptonight, ever.
My 7950 blew a mosfet too, the 4927N. But it was a fault of my own as I had "accidentally" placed the card on the metal frame of my case and switched on the power, and with no black plate, I am guessing that something shorted when the PCB contacts touched the metal frame. My PSU shut itself off, but not quick enough to save the 7950
Ordered a replacement part, taking it to a repair shop to get old one replaced. Hoping for the best.Apart from the Fan bug, I havent noticed any irregularaties with the mining. Though, one the fan shut down to 0% and I could smell that over heating/ burning smell from one of the 480's that had 0% fan and system rebooted. Everything was well afterwards. Luckily.
Also thanks to Wolf0 and Zanit(?) for sharing the miner and the config. I am mining at around 730 on 480's. With the :-
Threads 2
Rawintensity 504
Worksize 4
Drawing a total watt of around 360W for 3x 480's + system.
If anyone has any other settings please do share.
I am trying to improve the performance for CL Kernel and it's calling.
and, I have tried to compile, but I have failed,
can I get an idea about the compiling environments ?
I have tried
(1) install MinGW
(2) install All Packages
(3) compiled and installed jansson-2.8
(4) copied, AMD_SDK 3.0 x86 (bin/lib/include)
(5) I have tried to compile using Makefile it can not compile
it has generate an error : inconsistent operand constranint in an 'asm'
so, temporary, I have disabled
void cryptonight_hash_aesni in cryptonight.c & main.c
(6) I got so many errors,
so, I have added
-lws2_32
(7) But, some error has not been cleared,, Could you give me an advice ?
and, I have tried to compile, but I have failed,
can I get an idea about the compiling environments ?
I have tried
(1) install MinGW
(2) install All Packages
(3) compiled and installed jansson-2.8
(4) copied, AMD_SDK 3.0 x86 (bin/lib/include)
(5) I have tried to compile using Makefile it can not compile
it has generate an error : inconsistent operand constranint in an 'asm'
so, temporary, I have disabled
void cryptonight_hash_aesni in cryptonight.c & main.c
(6) I got so many errors,
so, I have added
-lws2_32
(7) But, some error has not been cleared,, Could you give me an advice ?
Quote
net.o:net.c:(.text+0x97): undefined reference to `getaddrinfo'
net.o:net.c:(.text+0x121): undefined reference to `freeaddrinfo'
net.o:net.c:(.text+0x14c): undefined reference to `freeaddrinfo'
collect2.exe: error: ld returned 1 exit status
make: *** [all] Error 1
Administrator@ENG ~/xmr
$
net.o:net.c:(.text+0x121): undefined reference to `freeaddrinfo'
net.o:net.c:(.text+0x14c): undefined reference to `freeaddrinfo'
collect2.exe: error: ld returned 1 exit status
make: *** [all] Error 1
Administrator@ENG ~/xmr
$
Oh, my HD7950 just about burst into flames! Minimum fan speed is 60% for that old dog. I'm guessing the number of transactions per second has skyrocketed, are we under attack or just really popular?
edit:
I can smell piney resin enamel from the inductor windings vaporizing combined with your classic electrical fire odor. Yeah that puppy just got hot...
edit2:
The RX480 seems to be drawing considerably more than the manufacturers claimed 110W, I can now smell a hint of pine coming from my AX1200 PSU.... The GPU is however chilling.
edit3:
Insane, I had to shut the HD7950 down, the miner's AMD dual core CPU and PSU started bellowing toxic fumes. Both too hot to touch. I feel queezy.
WTF LOL
I'll try again later after it cools down and smell dissipates with CPU mining DISABLED!!! WOW I CAN"T BELIEVE THAT!!!
edit4:
Mining XMR is NOT for the faint of heart. Nobody is going to crack cryptonight, ever.
edit:
I can smell piney resin enamel from the inductor windings vaporizing combined with your classic electrical fire odor. Yeah that puppy just got hot...
edit2:
The RX480 seems to be drawing considerably more than the manufacturers claimed 110W, I can now smell a hint of pine coming from my AX1200 PSU.... The GPU is however chilling.
edit3:
Insane, I had to shut the HD7950 down, the miner's AMD dual core CPU and PSU started bellowing toxic fumes. Both too hot to touch. I feel queezy.
WTF LOL
I'll try again later after it cools down and smell dissipates with CPU mining DISABLED!!! WOW I CAN"T BELIEVE THAT!!!
edit4:
Mining XMR is NOT for the faint of heart. Nobody is going to crack cryptonight, ever.
Ah! That clears it up - thanks Wolf0!
Hey Wolf0 - I've been looking through the github source and using that as an example to work through some OpenCL programming. I'm confused by something in the cn0 kernel:
__attribute__((reqd_work_group_size(WORKSIZE, 8, 1)))
__kernel void cn0(__global ulong *input, __global uint4 *Scratchpad, __global ulong *states)
{
...
Scratchpad += ((get_global_id(0) - get_global_offset(0)));
states += (25 * (get_global_id(0) - get_global_offset(0)));
...
So, the states pointer is to a buffer of 200 bytes per global threads. Makes sense - each work item offsets the pointer by 25 ulongs (64-bits per ulong or 8 bytes), for a total of 200 bytes per work item. However, the scratchpad is 2MB per global thread, but the pointer arithmetic just seems to advance it by one uint4 (4 x 4 bytes or 16 bytes). I would have expected to see something like:
Scratchpad += (131072 * (get_global_id(0) - get_global_offset(0)));
In other words, each work item should be offset by the full size of the 2MB buffer (131072 unit4's). What am I missing here?
__attribute__((reqd_work_group_size(WORKSIZE, 8, 1)))
__kernel void cn0(__global ulong *input, __global uint4 *Scratchpad, __global ulong *states)
{
...
Scratchpad += ((get_global_id(0) - get_global_offset(0)));
states += (25 * (get_global_id(0) - get_global_offset(0)));
...
So, the states pointer is to a buffer of 200 bytes per global threads. Makes sense - each work item offsets the pointer by 25 ulongs (64-bits per ulong or 8 bytes), for a total of 200 bytes per work item. However, the scratchpad is 2MB per global thread, but the pointer arithmetic just seems to advance it by one uint4 (4 x 4 bytes or 16 bytes). I would have expected to see something like:
Scratchpad += (131072 * (get_global_id(0) - get_global_offset(0)));
In other words, each work item should be offset by the full size of the 2MB buffer (131072 unit4's). What am I missing here?
Can someone explain to me how to lower the % of GPU being used? My computer is very slow when mining and would like it to be somewhat functional.
Would really appreciate some help.
Thanks
Would really appreciate some help.
Thanks
GPU utilization shouldn't make your computer slow unless the applications you are running need the GPU (games, for example).
By default, the xmr.conf enables CPU mining as well, which is generally very inefficient depending on your CPU. If you disable CPU mining in the xmr.conf, then your computer will run faster and will use much less energy and create less heat.
To disable CPU mining, set the "threads" value to 0 (zero) for the "-1" device, like this:
Quote
"index": -1,
"threads": 0,
"rawintensity": 16,
"worksize": 16
"threads": 0,
"rawintensity": 16,
"worksize": 16
Hello all!
New here, but not new to mining. I was a relatively early miner of bitcoin and I did fairly well with that a couple years ago.
Thanks to WolfO for this miner. I have recently complied it and got it running on Ubuntu 16.04 with AMD RX 470 using AMDGPU-Pro 16.50 driver.
It's working pretty well so far:
I am considering buying a couple more cards and putting them in other systems in my house (daughters' computers) and I obviously I want to mine with those too.
Some pools seem to allow you to differentiate between individual miners so that you can see separate stats for each one. Dwarfpool, for example, states:
Would this work with Wolf's GPU Miner? Would I just change the xmr.conf file and append a worker-ID number after my wallet address, like:
Also, does this work with all pools, or just with ones that specifically support it?
Thanks in advance for any help.
New here, but not new to mining. I was a relatively early miner of bitcoin and I did fairly well with that a couple years ago.
Thanks to WolfO for this miner. I have recently complied it and got it running on Ubuntu 16.04 with AMD RX 470 using AMDGPU-Pro 16.50 driver.
It's working pretty well so far:
Quote
[12:31:34] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.75H/s
[12:31:38] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.53H/s
[12:31:41] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.84H/s
[12:31:45] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.95H/s
[12:31:48] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.79H/s
[12:31:52] Thread 0, GPU ID 0, GPU Type: Ellesmere: 557.15H/s
[12:31:55] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.51H/s
[12:31:59] Thread 0, GPU ID 0, GPU Type: Ellesmere: 555.99H/s
[12:32:02] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.08H/s
[12:32:04] Share accepted: 60/60 (100.00%)
[12:32:04] Total Hashrate: 556.08H/s
[12:31:38] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.53H/s
[12:31:41] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.84H/s
[12:31:45] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.95H/s
[12:31:48] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.79H/s
[12:31:52] Thread 0, GPU ID 0, GPU Type: Ellesmere: 557.15H/s
[12:31:55] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.51H/s
[12:31:59] Thread 0, GPU ID 0, GPU Type: Ellesmere: 555.99H/s
[12:32:02] Thread 0, GPU ID 0, GPU Type: Ellesmere: 556.08H/s
[12:32:04] Share accepted: 60/60 (100.00%)
[12:32:04] Total Hashrate: 556.08H/s
I am considering buying a couple more cards and putting them in other systems in my house (daughters' computers) and I obviously I want to mine with those too.
Some pools seem to allow you to differentiate between individual miners so that you can see separate stats for each one. Dwarfpool, for example, states:
Quote
"You can determine your rigs separately for better vardiff, lower rejects and detailed statistic. Use workername: WALLET . WORKER-ID"
Would this work with Wolf's GPU Miner? Would I just change the xmr.conf file and append a worker-ID number after my wallet address, like:
Quote
"user": "47HpEzStPJGjUyRbj6DXJyN993s3busRLjVxxxxxxxxxxxxxxxxxxxxxxFgvQNizy9urPytC4tJB6sf yeuUKk1qZy.123",
Also, does this work with all pools, or just with ones that specifically support it?
Thanks in advance for any help.
Hi,
I have 3 rigs, 2 x 6xRX480-8gb (H97 anniversary, AMD driver 16.9.2 and 16.11.4) and 1 x 2xRX470 (B85M-G43, AMD driver 16.11.4).
With the 2xRX470 rig, the Wolf's miner works fine.
1) with claymore's I have 1380h/s, with wolf's 1425h/s ( Thanks Za1n for config)
2) Power on the wall 330w --> 290w
Thanks Wolf.
All rigs use Windows 8.1 Embedded Industry Pro, same AMD drivers (tested : 16.9.2 and 16.11.4) but ... same issue with my 2 x 6xRX480-8gb rigs : Error -1 when calling clGetDeviceIDs for number of devices.
I used 6xgpu_mod for 2 x 6xRX480-8gb rigs ( https://lbr.id.lv/6xgpu_mod/6xGPU_mod_download.html)
I have 3 rigs, 2 x 6xRX480-8gb (H97 anniversary, AMD driver 16.9.2 and 16.11.4) and 1 x 2xRX470 (B85M-G43, AMD driver 16.11.4).
With the 2xRX470 rig, the Wolf's miner works fine.
1) with claymore's I have 1380h/s, with wolf's 1425h/s ( Thanks Za1n for config)
2) Power on the wall 330w --> 290w
Thanks Wolf.
All rigs use Windows 8.1 Embedded Industry Pro, same AMD drivers (tested : 16.9.2 and 16.11.4) but ... same issue with my 2 x 6xRX480-8gb rigs : Error -1 when calling clGetDeviceIDs for number of devices.
I used 6xgpu_mod for 2 x 6xRX480-8gb rigs ( https://lbr.id.lv/6xgpu_mod/6xGPU_mod_download.html)
Do the rigs with issues have on-board nvidia GPUs on the mainboards? I don't recall any errors but the nvidia GPU in the desktop is preventing me from running the wolf miner.
H97 anniversary motherboard support only Intel® HD Graphics Built-in, no on-board GPU. I have G1840 or G1820 CPUs with intel HD (Haswell) GPU like my 2xRX470.
Hi,
I have 3 rigs, 2 x 6xRX480-8gb (H97 anniversary, AMD driver 16.9.2 and 16.11.4) and 1 x 2xRX470 (B85M-G43, AMD driver 16.11.4).
With the 2xRX470 rig, the Wolf's miner works fine.
1) with claymore's I have 1380h/s, with wolf's 1425h/s ( Thanks Za1n for config)
2) Power on the wall 330w --> 290w
Thanks Wolf.
All rigs use Windows 8.1 Embedded Industry Pro, same AMD drivers (tested : 16.9.2 and 16.11.4) but ... same issue with my 2 x 6xRX480-8gb rigs : Error -1 when calling clGetDeviceIDs for number of devices.
I used 6xgpu_mod for 2 x 6xRX480-8gb rigs ( https://lbr.id.lv/6xgpu_mod/6xGPU_mod_download.html)
I have 3 rigs, 2 x 6xRX480-8gb (H97 anniversary, AMD driver 16.9.2 and 16.11.4) and 1 x 2xRX470 (B85M-G43, AMD driver 16.11.4).
With the 2xRX470 rig, the Wolf's miner works fine.
1) with claymore's I have 1380h/s, with wolf's 1425h/s ( Thanks Za1n for config)
2) Power on the wall 330w --> 290w
Thanks Wolf.
All rigs use Windows 8.1 Embedded Industry Pro, same AMD drivers (tested : 16.9.2 and 16.11.4) but ... same issue with my 2 x 6xRX480-8gb rigs : Error -1 when calling clGetDeviceIDs for number of devices.
I used 6xgpu_mod for 2 x 6xRX480-8gb rigs ( https://lbr.id.lv/6xgpu_mod/6xGPU_mod_download.html)
Do the rigs with issues have on-board nvidia GPUs on the mainboards? I don't recall any errors but the nvidia GPU in the desktop is preventing me from running the wolf miner.
Hi,
I have 3 rigs, 2 x 6xRX480-8gb (H97 anniversary, AMD driver 16.9.2 and 16.11.4) and 1 x 2xRX470 (B85M-G43, AMD driver 16.11.4).
With the 2xRX470 rig, the Wolf's miner works fine.
1) with claymore's I have 1380h/s, with wolf's 1425h/s ( Thanks Za1n for config)
2) Power on the wall 330w --> 290w
Thanks Wolf.
All rigs use Windows 8.1 Embedded Industry Pro, same AMD drivers (tested : 16.9.2 and 16.11.4) but ... same issue with my 2 x 6xRX480-8gb rigs : Error -1 when calling clGetDeviceIDs for number of devices.
I used 6xgpu_mod for 2 x 6xRX480-8gb rigs ( https://lbr.id.lv/6xgpu_mod/6xGPU_mod_download.html)
I have 3 rigs, 2 x 6xRX480-8gb (H97 anniversary, AMD driver 16.9.2 and 16.11.4) and 1 x 2xRX470 (B85M-G43, AMD driver 16.11.4).
With the 2xRX470 rig, the Wolf's miner works fine.
1) with claymore's I have 1380h/s, with wolf's 1425h/s ( Thanks Za1n for config)
2) Power on the wall 330w --> 290w
Thanks Wolf.
All rigs use Windows 8.1 Embedded Industry Pro, same AMD drivers (tested : 16.9.2 and 16.11.4) but ... same issue with my 2 x 6xRX480-8gb rigs : Error -1 when calling clGetDeviceIDs for number of devices.
I used 6xgpu_mod for 2 x 6xRX480-8gb rigs ( https://lbr.id.lv/6xgpu_mod/6xGPU_mod_download.html)
I get error on all of my rigs win 10 64bit i tried different drivers but still the same result, anyone has solution? I have 5gpus on 3 rigs and 7gpus on rest. For example all runs peachy for weeks on ZEC
Thanks in advance
Error -1 when calling clGetDeviceIDs for number of devices
Thanks in advance
Error -1 when calling clGetDeviceIDs for number of devices
Well I don't know what peachy is or ZEC but, it sounds like your miner is asking the Open CL drivers how many cards you have but it can't because the CL drivers are pooched somehow.
I would download the latest Radeon drivers from AMD and also the AMD APP SDK package, chances are that will clear it up. If not it couldn't hurt to update your drivers in any case.
Jump to: