I really can't get it right with that card.
I think this voltage is linked to memory voltage.
I'm on 19.8.1 now, and it doesn't get lower than 950 mv.
I played around with PPT reg files and both cards couldn't do more than 500 Mhz memory...
If I set 1000 mv in P4 memory setting, the core gets to 1000 mv also.
The other (MSI) is playing well @ 1200/825 mv, that's great.
I really can't figure out this crazyness in Vega's voltage/speed control. It drives me crazy.
Topic: [ANN] TeamRedMiner v0.10.10 - Ironfish/Kaspa/ZIL/Kawpow/Etchash and More - page 63. (Read 211877 times)
August 13, 2019, 04:35:08 PM
August 13, 2019, 02:30:39 AM
No, I didn't change ppt table or did any reg hack.
I'm happy with the hasrate. I prefer 1200 Mhz @ 800 mv than 1450 Mhz @ 866 mv.
I can try higher clock speeds later.
But for now I really can't figure why core voltage setting sticks on one card, but ir doesn't on the other.
I'm happy with the hasrate. I prefer 1200 Mhz @ 800 mv than 1450 Mhz @ 866 mv.
I can try higher clock speeds later.
But for now I really can't figure why core voltage setting sticks on one card, but ir doesn't on the other.
Hi! What driver version are you running, scrolled back but couldn't find it? I'm guessing the driver might be stubborn with the monitor GPU. I actually don't think I've tested the miner at all on Vegas driving monitor(s) under the newer Adrenaline drivers. Also, have you tried disabling p7 and running in a lower p-state? Just curious if that mechanism for pushing down clocks/voltage works.
For your AMD memtweak issues, I'd start with a Administrator command prompt, not a .bat file allowed to elevate privileges. Just find Command prompt in the start menu, right click and choose "Run as Administrator". Then, cd to your dir with the AMD mem tweak and run if from that prompt. If it works, it's something with the rest of your .bat file setup. If you still get the missing EIO.dll issue, well, is the EIO.dll file really stored in the same dir then?
. Driver is 19.7.1.
August 12, 2019, 04:34:51 PM
No, I didn't change ppt table or did any reg hack.
I'm happy with the hasrate. I prefer 1200 Mhz @ 800 mv than 1450 Mhz @ 866 mv.
I can try higher clock speeds later.
But for now I really can't figure why core voltage setting sticks on one card, but ir doesn't on the other.
I'm happy with the hasrate. I prefer 1200 Mhz @ 800 mv than 1450 Mhz @ 866 mv.
I can try higher clock speeds later.
But for now I really can't figure why core voltage setting sticks on one card, but ir doesn't on the other.
Hi! What driver version are you running, scrolled back but couldn't find it? I'm guessing the driver might be stubborn with the monitor GPU. I actually don't think I've tested the miner at all on Vegas driving monitor(s) under the newer Adrenaline drivers. Also, have you tried disabling p7 and running in a lower p-state? Just curious if that mechanism for pushing down clocks/voltage works.
For your AMD memtweak issues, I'd start with a Administrator command prompt, not a .bat file allowed to elevate privileges. Just find Command prompt in the start menu, right click and choose "Run as Administrator". Then, cd to your dir with the AMD mem tweak and run if from that prompt. If it works, it's something with the rest of your .bat file setup. If you still get the missing EIO.dll issue, well, is the EIO.dll file really stored in the same dir then?
. August 12, 2019, 03:21:33 PM
No, I didn't change ppt table or did any reg hack.
I'm happy with the hasrate. I prefer 1200 Mhz @ 800 mv than 1450 Mhz @ 866 mv.
I can try higher clock speeds later.
But for now I really can't figure why core voltage setting sticks on one card, but ir doesn't on the other.
I'm happy with the hasrate. I prefer 1200 Mhz @ 800 mv than 1450 Mhz @ 866 mv.
I can try higher clock speeds later.
But for now I really can't figure why core voltage setting sticks on one card, but ir doesn't on the other.
August 12, 2019, 03:08:13 PM
Samsung have 2300+HR on 860/870mv
First you must use 64 bios and your mem go on 1100mhz.
my vega 56 samsung have 2360hr and work on 1100mem/866mv and gpu on 1440/866mv
you chg your PPT table?
ODT for samsung:
Name=max1
GPU_P0=852;800;0
GPU_P1=991;801;0
GPU_P2=1084;802;0
GPU_P3=1100;803;0
GPU_P4=1125;804;0
GPU_P5=1200;864
GPU_P6=1400;865
GPU_P7=1450;866
Mem_P0=167;800;0
Mem_P1=500;800;0
Mem_P2=800;802
Mem_P3=1100;866
Mem_TimingLevel=0
Fan_P0=56;45
Fan_P1=58;51
Fan_P2=60;60
Fan_P3=62;70
Fan_P4=64;80
Fan_ZeroRPM=0
Fan_Acoustic=1800
Power_Target=0
ODT 0.2.8 wont new drive if you use 1..86 new ODT not work good
your gpu1 use to much power. try reinstal driver or chg ppt table for this gpu
your gpu0 use ultra low power:) this is super
First you must use 64 bios and your mem go on 1100mhz.
my vega 56 samsung have 2360hr and work on 1100mem/866mv and gpu on 1440/866mv
you chg your PPT table?
ODT for samsung:
Name=max1
GPU_P0=852;800;0
GPU_P1=991;801;0
GPU_P2=1084;802;0
GPU_P3=1100;803;0
GPU_P4=1125;804;0
GPU_P5=1200;864
GPU_P6=1400;865
GPU_P7=1450;866
Mem_P0=167;800;0
Mem_P1=500;800;0
Mem_P2=800;802
Mem_P3=1100;866
Mem_TimingLevel=0
Fan_P0=56;45
Fan_P1=58;51
Fan_P2=60;60
Fan_P3=62;70
Fan_P4=64;80
Fan_ZeroRPM=0
Fan_Acoustic=1800
Power_Target=0
ODT 0.2.8 wont new drive if you use 1..86 new ODT not work good
your gpu1 use to much power. try reinstal driver or chg ppt table for this gpu
your gpu0 use ultra low power:) this is super
August 12, 2019, 03:03:21 PM
Here:
It's the Sapphire Pulse where the voltages settings doesn't stick. It's driving the monitor.
Tried ODT 0.2.8 and it's the same. Not really sure what's hapenning.
It's the Sapphire Pulse where the voltages settings doesn't stick. It's driving the monitor.
Tried ODT 0.2.8 and it's the same. Not really sure what's hapenning.
August 12, 2019, 02:49:28 PM
I'm having some issues, though.
I can't change memory settings with CLI, only GUI version.
Making a bat file and giving permission to run as administrator to WinAMDTweak.exe does nothing to memory timings, running from an administrator command prompt gives something about a missing or can't open EIO.DLL file.
GUI does work, just more tedious to load values and apply to 2 cards.
2x Vega 56 (MSI Airboost and Sapphire Pulse) both do 2050 h/s @ 1200 core / 940 memory, that's nothing short of amazing.
However, I'm don't know if I'm failing to set voltages properly. Power consumption went down 50W applying these voltage settings (same for both cards), HWinfo shows one card @0.8 vcore and other @0.925 vcore.
Any ideas?
I can't change memory settings with CLI, only GUI version.
Making a bat file and giving permission to run as administrator to WinAMDTweak.exe does nothing to memory timings, running from an administrator command prompt gives something about a missing or can't open EIO.DLL file.
GUI does work, just more tedious to load values and apply to 2 cards.
2x Vega 56 (MSI Airboost and Sapphire Pulse) both do 2050 h/s @ 1200 core / 940 memory, that's nothing short of amazing.
However, I'm don't know if I'm failing to set voltages properly. Power consumption went down 50W applying these voltage settings (same for both cards), HWinfo shows one card @0.8 vcore and other @0.925 vcore.
Any ideas?
if you see EIO.dll error you must restart comp adn not start srbminer. srbminer block amd mem tweak if not use --disabletweaking.
if dont work after restart you must go dell folder CACHE in srbminer folder.
Your card have hynix mem=?
hynix on CNR have 2100+ HR
Try this setings and chg mem clock after test:
6*hynix-12500+HR - 1050W
WinAMDTweak.exe --RAS 24 --rp 14 --rc 35 --rfc 168 --rrds 4 --rrdl 5 --rcdrd 19 --rcdwr 7 --REF 15600
teamredminer.exe -a cnr -o stratum+tcp://monero.miner.rocks:5555 -u ADRES -p Livada2 -d 0,1 --cn_config=14*14:BAB,14*14:BAB --bus_reorder --watchdog_script=restart-cnr.bat
ODT:
Name=CNR5
GPU_P0=852;800;0
GPU_P1=991;801;0
GPU_P2=1084;802;0
GPU_P3=1100;803;0
GPU_P4=1125;804;0
GPU_P5=1150;805;0
GPU_P6=1350;815;0
GPU_P7=1360;840
Mem_P0=167;800;0
Mem_P1=500;800;0
Mem_P2=800;801;0
Mem_P3=929;860
Fan_Min=1700
Fan_Max=3300
Fan_Target=56
Fan_Acoustic=2400
Power_Temp=70
Power_Target=0
My cards have Samsung memory. I'm happy with 2050 @ 1200 Mhz core, no point in having it @ ~1400 for 50 h/s more.
I rebooted / load defaults and didn't run srb miner.
But I actually prefer the GUI mode because it can handle all, timings and P states.
My problem is with undervolting. One card seems to run fine @ 825 mv core, reported in HWinfo, power consumption went down, but the other is still @ 925 mv after applied same P states to it.
That's where I need help.
August 12, 2019, 02:29:48 PM
I'm having some issues, though.
I can't change memory settings with CLI, only GUI version.
Making a bat file and giving permission to run as administrator to WinAMDTweak.exe does nothing to memory timings, running from an administrator command prompt gives something about a missing or can't open EIO.DLL file.
GUI does work, just more tedious to load values and apply to 2 cards.
2x Vega 56 (MSI Airboost and Sapphire Pulse) both do 2050 h/s @ 1200 core / 940 memory, that's nothing short of amazing.
However, I'm don't know if I'm failing to set voltages properly. Power consumption went down 50W applying these voltage settings (same for both cards), HWinfo shows one card @0.8 vcore and other @0.925 vcore.
Any ideas?
I can't change memory settings with CLI, only GUI version.
Making a bat file and giving permission to run as administrator to WinAMDTweak.exe does nothing to memory timings, running from an administrator command prompt gives something about a missing or can't open EIO.DLL file.
GUI does work, just more tedious to load values and apply to 2 cards.
2x Vega 56 (MSI Airboost and Sapphire Pulse) both do 2050 h/s @ 1200 core / 940 memory, that's nothing short of amazing.
However, I'm don't know if I'm failing to set voltages properly. Power consumption went down 50W applying these voltage settings (same for both cards), HWinfo shows one card @0.8 vcore and other @0.925 vcore.
Any ideas?
if you see EIO.dll error you must restart comp adn not start srbminer. srbminer block amd mem tweak if not use --disabletweaking.
if dont work after restart you must go dell folder CACHE in srbminer folder.
Your card have hynix mem=?
hynix on CNR have 2100+ HR
Try this setings and chg mem clock after test:
6*hynix-12500+HR - 1050W
WinAMDTweak.exe --RAS 24 --rp 14 --rc 35 --rfc 168 --rrds 4 --rrdl 5 --rcdrd 19 --rcdwr 7 --REF 15600
teamredminer.exe -a cnr -o stratum+tcp://monero.miner.rocks:5555 -u ADRES -p Livada2 -d 0,1 --cn_config=14*14:BAB,14*14:BAB --bus_reorder --watchdog_script=restart-cnr.bat
ODT:
Name=CNR5
GPU_P0=852;800;0
GPU_P1=991;801;0
GPU_P2=1084;802;0
GPU_P3=1100;803;0
GPU_P4=1125;804;0
GPU_P5=1150;805;0
GPU_P6=1350;815;0
GPU_P7=1360;840
Mem_P0=167;800;0
Mem_P1=500;800;0
Mem_P2=800;801;0
Mem_P3=929;860
Fan_Min=1700
Fan_Max=3300
Fan_Target=56
Fan_Acoustic=2400
Power_Temp=70
Power_Target=0
August 12, 2019, 10:01:11 AM
I'm having some issues, though.
I can't change memory settings with CLI, only GUI version.
Making a bat file and giving permission to run as administrator to WinAMDTweak.exe does nothing to memory timings, running from an administrator command prompt gives something about a missing or can't open EIO.DLL file.
GUI does work, just more tedious to load values and apply to 2 cards.
2x Vega 56 (MSI Airboost and Sapphire Pulse) both do 2050 h/s @ 1200 core / 940 memory, that's nothing short of amazing.
However, I'm don't know if I'm failing to set voltages properly. Power consumption went down 50W applying these voltage settings (same for both cards), HWinfo shows one card @0.8 vcore and other @0.925 vcore.
Any ideas?
I can't change memory settings with CLI, only GUI version.
Making a bat file and giving permission to run as administrator to WinAMDTweak.exe does nothing to memory timings, running from an administrator command prompt gives something about a missing or can't open EIO.DLL file.
GUI does work, just more tedious to load values and apply to 2 cards.
2x Vega 56 (MSI Airboost and Sapphire Pulse) both do 2050 h/s @ 1200 core / 940 memory, that's nothing short of amazing.
However, I'm don't know if I'm failing to set voltages properly. Power consumption went down 50W applying these voltage settings (same for both cards), HWinfo shows one card @0.8 vcore and other @0.925 vcore.
Code:
-
-
-
-
Any ideas?
August 11, 2019, 03:24:38 PM
Thought this was a shaddy miner from the spam on SRB thread, but now I see it's a very serious project, and it does deliver. Congratulations.
Thanks, and yes, I noted the drama in the SRB thread yesterday. Needless to say perhaps, no one in this project is happy with how TRM got thrown into that debate, and we're not affiliated in any way with the person(s) involved. I'm glad it seems to have quieted down. I do have a strict rule myself of not posting in other miners' threads as well, it's a very slippery slope once you cross that bridge.
There are pros and cons with all miners, across all the available CN miners we have probably spent the most time of all on our kernels, handrolled from scratch in GCN assembly, obsessing about every single little instruction involved. I believe it shows in hashrate and power draw in the vast majority of all setups. That said, we're instead lagging in other feature areas, like http monitoring, pool failover, etc, we tend to move on to the next algo instead. Furthermore, some people can't picture using anything but open source miners, other get a knee jerk reaction from the relatively high dev fee of 2.5%. If they knew how hard we worked on this project, or the time it takes to roll out these kernels, maybe it would change their minds, but I won't judge anyone or complain even though I'm pretty sure they could earn a few more cents/day/gpu, it's up to the users to choose their miner.
August 11, 2019, 12:07:55 PM
Thought this was a shaddy miner from the spam on SRB thread, but now I see it's a very serious project, and it does deliver. Congratulations.
August 10, 2019, 02:17:05 AM
Team Red Miner v0.5.7 released
https://github.com/todxx/teamredminer/releases
Changes in v0.5.7
This release brings a couple new algos: CN conceal and cuckarood29!
CN conceal is a bit picky about memory timings, so please checkout the CN_MAX_YOU_VEGA.txt doc before switching your rigs to it.
https://github.com/todxx/teamredminer/releases
Changes in v0.5.7
- Added CN conceal algo for Conceal (CCX).
- Added cuckarood29 algo for grin.
This release brings a couple new algos: CN conceal and cuckarood29!
CN conceal is a bit picky about memory timings, so please checkout the CN_MAX_YOU_VEGA.txt doc before switching your rigs to it.
I am unable to run cn_conceal algo. The miner always fail GPU initialization beyond the first GPU. How to fix this?
It sounds like you can run other CN variants? Conceal has the exact same init process as all other 2MB variants, including loading the same binary, so it’d be very strange if it got stuck at init
. Can you run the miner with —debug and pm the logs/output please? I’m
August 09, 2019, 10:22:12 PM
Team Red Miner v0.5.7 released
https://github.com/todxx/teamredminer/releases
Changes in v0.5.7
This release brings a couple new algos: CN conceal and cuckarood29!
CN conceal is a bit picky about memory timings, so please checkout the CN_MAX_YOU_VEGA.txt doc before switching your rigs to it.
https://github.com/todxx/teamredminer/releases
Changes in v0.5.7
- Added CN conceal algo for Conceal (CCX).
- Added cuckarood29 algo for grin.
This release brings a couple new algos: CN conceal and cuckarood29!
CN conceal is a bit picky about memory timings, so please checkout the CN_MAX_YOU_VEGA.txt doc before switching your rigs to it.
I am unable to run cn_conceal algo. The miner always fail GPU initialization beyond the first GPU. How to fix this?
August 08, 2019, 04:28:44 AM
What is the best config for RX570 8GB and RX580 8GB?
In other miners i can reach 1200 hash on xhv per card, i cant reach 1080 hash in teamredminer.
In other miners i can reach 1200 hash on xhv per card, i cant reach 1080 hash in teamredminer.
So, what does the auto-tuning do for you when you start the miner without arguments?
Just ran a quick test on two 570+580 in my linux test rig, ended up using 16+14:AAA for 1257 h/s per gpu. Can't say I was using any clocks specifically tuned for CN, they were both running at coreclk 1200@875mV, memclk 2100@875mV. These are both Sapphire Nitro+ with Samsung mem, Ubermix 3.1 straps iirc (not 100% sure). No tREF mod / rx-boost applied.
August 08, 2019, 03:10:17 AM
What is the best config for RX570 8GB and RX580 8GB?
In other miners i can reach 1200 hash on xhv per card, i cant reach 1080 hash in teamredminer.
In other miners i can reach 1200 hash on xhv per card, i cant reach 1080 hash in teamredminer.
August 06, 2019, 10:56:06 AM
maybe help you for 5700 https://gpuopen.com/wp-content/uploads/2019/08/RDNA_Shader_ISA_7July2019.pdf
August 06, 2019, 08:35:15 AM
Team Red Miner v0.5.7 released
https://github.com/todxx/teamredminer/releases
Changes in v0.5.7
This release brings a couple new algos: CN conceal and cuckarood29!
CN conceal is a bit picky about memory timings, so please checkout the CN_MAX_YOU_VEGA.txt doc before switching your rigs to it.
https://github.com/todxx/teamredminer/releases
Changes in v0.5.7
- Added CN conceal algo for Conceal (CCX).
- Added cuckarood29 algo for grin.
This release brings a couple new algos: CN conceal and cuckarood29!
CN conceal is a bit picky about memory timings, so please checkout the CN_MAX_YOU_VEGA.txt doc before switching your rigs to it.
My 4 x Vega 64 are mining at 4.5kH/s each. Settings like on CNR. "Cool" algo in terms of power and temps.
August 06, 2019, 02:13:06 AM
how to run ccx what command line
what command lineHi qwep1,
You can run CN conceal the same way as cnr, except use "cn_conceal" as the algorithm instead of "cnr".
August 06, 2019, 01:09:16 AM
how to run ccx what command line
what command line August 06, 2019, 12:47:49 AM
Team Red Miner v0.5.7 released
https://github.com/todxx/teamredminer/releases
Changes in v0.5.7
This release brings a couple new algos: CN conceal and cuckarood29!
CN conceal is a bit picky about memory timings, so please checkout the CN_MAX_YOU_VEGA.txt doc before switching your rigs to it.
How much memory is needed for cuckarood29? https://github.com/todxx/teamredminer/releases
Changes in v0.5.7
- Added CN conceal algo for Conceal (CCX).
- Added cuckarood29 algo for grin.
This release brings a couple new algos: CN conceal and cuckarood29!
CN conceal is a bit picky about memory timings, so please checkout the CN_MAX_YOU_VEGA.txt doc before switching your rigs to it.
RX 570 4GB not work on Win7 (Blockchain drivers)
[2019-08-06 07:50:40] Auto-detected AMD OpenCL platform 0
[2019-08-06 07:50:43] Initializing GPU 0.
[2019-08-06 07:50:43] OpenCL enqueue error: -55
[2019-08-06 07:50:43] Warning: failed to initialize device number 0 (-28), retrying.
[2019-08-06 07:50:43] OpenCL enqueue error: -55
[2019-08-06 07:50:43] Warning: failed to initialize device number 0 (-28), retrying.
[2019-08-06 07:50:43] OpenCL enqueue error: -55
[2019-08-06 07:50:43] Warning: failed to initialize device number 0 (-28), retrying.
[2019-08-06 07:50:44] OpenCL enqueue error: -55
[2019-08-06 07:50:44] Warning: failed to initialize device number 0 (-28), retrying.
[2019-08-06 07:50:44] OpenCL enqueue error: -55
[2019-08-06 07:50:44] Warning: failed to initialize device number 0 (-28), retrying.
[2019-08-06 07:50:44] OpenCL enqueue error: -55
[2019-08-06 07:50:44] Warning: failed to initialize device number 0 (-28), retrying.
[2019-08-06 07:50:44] OpenCL enqueue error: -55
[2019-08-06 07:50:44] Warning: failed to initialize device number 0 (-28), retrying.
[2019-08-06 07:50:44] OpenCL enqueue error: -55
[2019-08-06 07:50:44] Warning: failed to initialize device number 0 (-28), retrying.
[2019-08-06 07:50:45] OpenCL enqueue error: -55
[2019-08-06 07:50:45] Warning: failed to initialize device number 0 (-28), retrying.
[2019-08-06 07:50:45] OpenCL enqueue error: -55
[2019-08-06 07:50:45] Warning: failed to initialize device number 0 (-28), retrying.
[2019-08-06 07:50:45] OpenCL enqueue error: -55
[2019-08-06 07:50:45] Failed to initialize device idx 0 (-28)
[2019-08-06 07:50:45] Successful clean shutdown.
My apologies, we should list somewhere that our c31 and c29 kernels currently only work on 8GB+ GPUs.
I thought that we had a much nicer message in place than the errors that were printed for you.
Jump to: