[ANN] Best sgminer so far for x25x/lyra2Z/phi2/allium/lyra2REv3/yescrypt on AMD

[Ladydoge]

Beta5f:  Error -54 enqueueing kernel search10 onto command queue. (clEnqueueNDRangeKernel)
GPU 0 failure, disabling!

[1713582210]

1713582210

[1713582210]

1713582210

[1713582210]

1713582210

[1713582210]

1713582210

[fancyIX]

Beta5f release. x22i over 25% faster than beta5e, lyra2rev2 over 4x faster.
https://github.com/fancyIX/sgminer-phi2-branch/releases

Lyra2rev2?  or lyra2z... confused there... didn't know you did lyra2rev2.  not that it matters cause lyra2rev2 is asic country now

Lyra2re2.
It's almost impossible to to 4x faster for lyra2Z.

[fancyIX]

Beta5f:  Error -54 enqueueing kernel search10 onto command queue. (clEnqueueNDRangeKernel)
GPU 0 failure, disabling!

Maybe you can try a lower -I number, like "-I 22". Also set "-g 1" may help.

[Ladydoge]

Beta5f:  Error -54 enqueueing kernel search10 onto command queue. (clEnqueueNDRangeKernel)
GPU 0 failure, disabling!

Maybe you can try a lower -I number, like "-I 22". Also set "-g 1" may help.

I use intensity 16 and -g 1 didn't help. Previous beta works fine.

Log doesn't clear anything up I think

[08:52:29] Started sgminer v5.6.1.3
[08:52:29] * using Jansson 2.11
[08:52:30] CL Platform vendor: Advanced Micro Devices, Inc.
[08:52:30] CL Platform name: AMD Accelerated Parallel Processing
[08:52:30] CL Platform version: OpenCL 2.1 AMD-APP (2580.6)
[08:52:30] Platform devices: 1
[08:52:30]    0   Tahiti
[08:52:30] Probing for an alive pool
[08:52:30] Testing pool.dapool.me
[08:52:30] Stratum extranonce subscribe for pool.dapool.me
[08:52:31] Stratum authorisation success for pool.dapool.me
[08:52:31] pool.dapool.me alive
[08:52:31] Startup GPU initialization... Using settings from pool pool.dapool.me.
[08:52:31] Startup Pool No = 0
[08:52:31] Init GPU thread 0 GPU 0 virtual GPU 0
[08:52:31] CL Platform vendor: Advanced Micro Devices, Inc.
[08:52:31] CL Platform name: AMD Accelerated Parallel Processing
[08:52:31] CL Platform version: OpenCL 2.1 AMD-APP (2580.6)
[08:52:31] Platform devices: 1
[08:52:31]    0   Tahiti
[08:52:31] List of devices:
[08:52:31]    0   Tahiti
[08:52:31] Selected 0: Tahiti
[08:52:31] Maximum work size for this GPU (0) is 256.
[08:52:31] Your GPU (#0) has 32 compute units, and all AMD cards in the 7 series or newer (GCN cards)       have 64 shaders per compute unit - this means it has 2048 shaders.
[08:52:31] Initialising kernel x22i.cl with nfactor 10, n 1024
[08:52:31] initCl() finished. Found Tahiti
[08:52:31] pool.dapool.me difficulty changed to 0.025
[08:52:31] New block: 00000000000730795947fe963006e80b18be11c8bc55eaad72f35b64cab07595... diff 7.81K
[08:52:33] Work update message received
[08:52:33] Error -54: Enqueueing kernel search10 onto command queue. (clEnqueueNDRangeKernel)
[08:52:33] GPU 0 failure, disabling!

[fancyIX]

Beta5f:  Error -54 enqueueing kernel search10 onto command queue. (clEnqueueNDRangeKernel)
GPU 0 failure, disabling!

Maybe you can try a lower -I number, like "-I 22". Also set "-g 1" may help.

I use intensity 16 and -g 1 didn't help. Previous beta works fine.

Log doesn't clear anything up I think

[08:52:29] Started sgminer v5.6.1.3
[08:52:29] * using Jansson 2.11
[08:52:30] CL Platform vendor: Advanced Micro Devices, Inc.
[08:52:30] CL Platform name: AMD Accelerated Parallel Processing
[08:52:30] CL Platform version: OpenCL 2.1 AMD-APP (2580.6)
[08:52:30] Platform devices: 1
[08:52:30]    0   Tahiti
[08:52:30] Probing for an alive pool
[08:52:30] Testing pool.dapool.me
[08:52:30] Stratum extranonce subscribe for pool.dapool.me
[08:52:31] Stratum authorisation success for pool.dapool.me
[08:52:31] pool.dapool.me alive
[08:52:31] Startup GPU initialization... Using settings from pool pool.dapool.me.
[08:52:31] Startup Pool No = 0
[08:52:31] Init GPU thread 0 GPU 0 virtual GPU 0
[08:52:31] CL Platform vendor: Advanced Micro Devices, Inc.
[08:52:31] CL Platform name: AMD Accelerated Parallel Processing
[08:52:31] CL Platform version: OpenCL 2.1 AMD-APP (2580.6)
[08:52:31] Platform devices: 1
[08:52:31]    0   Tahiti
[08:52:31] List of devices:
[08:52:31]    0   Tahiti
[08:52:31] Selected 0: Tahiti
[08:52:31] Maximum work size for this GPU (0) is 256.
[08:52:31] Your GPU (#0) has 32 compute units, and all AMD cards in the 7 series or newer (GCN cards)       have 64 shaders per compute unit - this means it has 2048 shaders.
[08:52:31] Initialising kernel x22i.cl with nfactor 10, n 1024
[08:52:31] initCl() finished. Found Tahiti
[08:52:31] pool.dapool.me difficulty changed to 0.025
[08:52:31] New block: 00000000000730795947fe963006e80b18be11c8bc55eaad72f35b64cab07595... diff 7.81K
[08:52:33] Work update message received
[08:52:33] Error -54: Enqueueing kernel search10 onto command queue. (clEnqueueNDRangeKernel)
[08:52:33] GPU 0 failure, disabling!

Hold on. Did you remove the old .bin file from there folder where "sgminer" is?

[Ladydoge]

I ran beta5f from a new folder, so no old bins. I also tried using a bin file made by beta5f with old beta and got the same error.

[fancyIX]

I ran beta5f from a new folder, so no old bins. I also tried using a bin file made by beta5f with old beta and got the same error.

Just found a "bug" in the code. I guess you are not using "-w 256", right?
in file kernel/x22i.cl, line 1115, change "256" to "WORKSIZE" may solve your problem.

[Ladydoge]

I ran beta5f from a new folder, so no old bins. I also tried using a bin file made by beta5f with old beta and got the same error.

Just found a "bug" in the code. I guess you are not using "-w 256", right?
in file kernel/x22i.cl, line 1115, change "256" to "WORKSIZE" may solve your problem.

Yes, I usually use -w 128. Can you copy the whole line here please so I can search for it and edit? I use notepad to edit stuff lol, so no line numbers there.
Tried using -w 256. It works but I get 1.4 mhs, same hashrate as in beta5e++

[fancyIX]

I ran beta5f from a new folder, so no old bins. I also tried using a bin file made by beta5f with old beta and got the same error.

Just found a "bug" in the code. I guess you are not using "-w 256", right?
in file kernel/x22i.cl, line 1115, change "256" to "WORKSIZE" may solve your problem.

Yes, I usually use -w 128. Can you copy the whole line here please so I can search for it and edit? I use notepad to edit stuff lol, so no line numbers there.
Tried using -w 256. It works but I get 1.4 mhs, same hashrate as in beta5e++

I just updated the release files. Can you download again and try the new files?

[Ladydoge]

I ran beta5f from a new folder, so no old bins. I also tried using a bin file made by beta5f with old beta and got the same error.

Just found a "bug" in the code. I guess you are not using "-w 256", right?
in file kernel/x22i.cl, line 1115, change "256" to "WORKSIZE" may solve your problem.

Yes, I usually use -w 128. Can you copy the whole line here please so I can search for it and edit? I use notepad to edit stuff lol, so no line numbers there.
Tried using -w 256. It works but I get 1.4 mhs, same hashrate as in beta5e++

I just updated the release files. Can you download again and try the new files?

Now it works fine with -w 128. No improvements in hashrate over previous beta for me. Seems that 1.4 mhs is the maximum hashrate I can get from such an old card with these settings:

[fancyIX]

I ran beta5f from a new folder, so no old bins. I also tried using a bin file made by beta5f with old beta and got the same error.

Just found a "bug" in the code. I guess you are not using "-w 256", right?
in file kernel/x22i.cl, line 1115, change "256" to "WORKSIZE" may solve your problem.

Yes, I usually use -w 128. Can you copy the whole line here please so I can search for it and edit? I use notepad to edit stuff lol, so no line numbers there.
Tried using -w 256. It works but I get 1.4 mhs, same hashrate as in beta5e++

I just updated the release files. Can you download again and try the new files?

Now it works fine with -w 128. No improvements in hashrate over previous beta for me. Seems that 1.4 mhs is the maximum hashrate I can get from such an old card with these settings:

Which driver are you using? The improvement I made for this new release is pretty driver dependent. I made some changes tuned for the driver I am using (almost the newest version). If you are using old driver there may not be improvement.

[Ladydoge]

I ran beta5f from a new folder, so no old bins. I also tried using a bin file made by beta5f with old beta and got the same error.

Just found a "bug" in the code. I guess you are not using "-w 256", right?
in file kernel/x22i.cl, line 1115, change "256" to "WORKSIZE" may solve your problem.

Yes, I usually use -w 128. Can you copy the whole line here please so I can search for it and edit? I use notepad to edit stuff lol, so no line numbers there.
Tried using -w 256. It works but I get 1.4 mhs, same hashrate as in beta5e++

I just updated the release files. Can you download again and try the new files?

Now it works fine with -w 128. No improvements in hashrate over previous beta for me. Seems that 1.4 mhs is the maximum hashrate I can get from such an old card with these settings:

Which driver are you using? The improvement I made for this new release is pretty driver dependent. I made some changes tuned for the driver I am using (almost the newest version). If you are using old driver there may not be improvement.

Adrenalin 18.5.1 on win7. Which driver would you recommend?

[fancyIX]

I ran beta5f from a new folder, so no old bins. I also tried using a bin file made by beta5f with old beta and got the same error.

Just found a "bug" in the code. I guess you are not using "-w 256", right?
in file kernel/x22i.cl, line 1115, change "256" to "WORKSIZE" may solve your problem.

Yes, I usually use -w 128. Can you copy the whole line here please so I can search for it and edit? I use notepad to edit stuff lol, so no line numbers there.
Tried using -w 256. It works but I get 1.4 mhs, same hashrate as in beta5e++

I just updated the release files. Can you download again and try the new files?

Now it works fine with -w 128. No improvements in hashrate over previous beta for me. Seems that 1.4 mhs is the maximum hashrate I can get from such an old card with these settings:

Which driver are you using? The improvement I made for this new release is pretty driver dependent. I made some changes tuned for the driver I am using (almost the newest version). If you are using old driver there may not be improvement.

Adrenalin 18.5.1 on win7. Which driver would you recommend?

That seems pretty new. I guess that's all it can do for your card. Maybe the performance can be improved in next release.
What's the hashrate of other miners, like zjass or wiildog?

[Ladydoge]

Zjazz and Wildrig don't support Tahiti.

[john1010]

I got a LIBEAY32.dll error when running it, by the way I am mining SUQA with x22i algo, is this miner supported it? But someone in discord shared this miner with our group channel, I am using rx 580 8GB, I hope someone may share his sample bat config here.. Anyway thank you OP for this thread.

[Sara93]

I got a LIBEAY32.dll error when running it, by the way I am mining SUQA with x22i algo, is this miner supported it? But someone in discord shared this miner with our group channel, I am using rx 580 8GB, I hope someone may share his sample bat config here.. Anyway thank you OP for this thread.

To fix your mistake, you need to install OpenSSL, you can download it in the same place where the miner is loaded. Next time be more attentive

[john1010]

Supports Windows and Linux. Supports windows driver and Linux AMDGPU Pro driver. Supports most cards. Comes with pre-built binary kernels to speed up Ellesmere and Vega cards.


https://github.com/fancyIX/sgminer-phi2-branch/releases

Estimated speed (beta5f):
phi2,     rx 580: 3750khs; vega 56: 6750khs
lyra2Z,  rx 580: 1950khs; vega 56: 3600khs
allium,   rx 580: 4500khs, vega 56: 8200khs
x22i,     rx 580: 4600khs, vega 56: 6900khs


I know that's far below the speed of tdxminer or teamredminer. Hope one day it can catch up.

Thanks for sharing your miner on SUQA's ANN, by the way im using it now, my problem is the latest version of afterburner only support 8cards, so i have no option to set up my fan speed and also the power.. The old SGminer has the capacity to adjust the fan, power and also the mem core of the card, My question is, is this miner is capable of doing that in the config .bat file?

[john1010]

I got a LIBEAY32.dll error when running it, by the way I am mining SUQA with x22i algo, is this miner supported it? But someone in discord shared this miner with our group channel, I am using rx 580 8GB, I hope someone may share his sample bat config here.. Anyway thank you OP for this thread.

To fix your mistake, you need to install OpenSSL, you can download it in the same place where the miner is loaded. Next time be more attentive

Don't preach me brother.. Of course i have that on my system.. My concern here is the sample bat file to control my card.. Anyway thanks for mentioned.    and that error is from my personal pc only I tried to run the miner without adjusting my VM, I'm successfully run the miner on my rig now.

[Sara93]

Miner 5.6.1.3.b5f driver Adrenalin 18.6.1 RX 470 4GB 1100/1900
x22i 3.1 Mhs
phi2 2.6 Mhs (no change)
lyra2Z 1.4 Mhs (no change)
allium 3.0 Mhs (no change)
lyra2rev2 22.7 Mhs
P.S. On the x22i algorithm on the same cards (in the same settings) WildRig Multi 0.12.7 gives 3.3 Mhs. Excellent result !!!

[Ladydoge]

r9 280x 15 mhs on lyra2rev2. If only this miner had been released when this algo wasn't asic ridden