kerney666
Member
Offline
Activity: 658
Merit: 86
|
|
December 20, 2018, 03:09:37 AM |
|
Just a small update on how my small farm is doing, I been running this miner for a week now pretty stable with almost no issues. Not sure how I can check for the percentage of rejected and accepted shares, but all the numbers seem to have evened out. All my miners have been running almost all the same across the board. Maybe a few points of difference between them but they can all hit at least 1900 sol/s for my vega 56's.
A few days ago I read a reddit post of a dude saying he was able to hit 2000 sol/s on his vega 56 with an extremely reduced power draw. He was using the recently released adrenalin drivers but didn't give out much more information. So far I have downloaded the second to latest drivers and have been able to closely replicate the results i had with using the soft powerplay table mod and using overdriveN tool. It does feel like there is room for fine tuning but i wouldn't know where to start. Ideally I would like to hit the 2000 sol/s using 115w on gpu-z. Maybe not that low but having a system total power draw of 500w or under for 3 vega 56 would be very nice.
Hi! There are a number of skilled Vega tuners in the forum, some also hang out in the supportxmr chatango chat room (pbfarmer, derp4vega, saltednutz, etc). I know derp4vega has been kind enough to share his ppt mods earlier, but of course it's entirely up to him. Personally, I've done a fair amount of mining on my own Vega rigs, but I'm nowhere near these guys in experience with ppt mods.
|
|
|
|
pbfarmer
Member
Offline
Activity: 340
Merit: 29
|
|
December 20, 2018, 10:04:21 PM Last edit: December 20, 2018, 10:21:20 PM by pbfarmer Merited by kerney666 (1), todxx (1) |
|
Just a small update on how my small farm is doing, I been running this miner for a week now pretty stable with almost no issues. Not sure how I can check for the percentage of rejected and accepted shares, but all the numbers seem to have evened out. All my miners have been running almost all the same across the board. Maybe a few points of difference between them but they can all hit at least 1900 sol/s for my vega 56's.
A few days ago I read a reddit post of a dude saying he was able to hit 2000 sol/s on his vega 56 with an extremely reduced power draw. He was using the recently released adrenalin drivers but didn't give out much more information. So far I have downloaded the second to latest drivers and have been able to closely replicate the results i had with using the soft powerplay table mod and using overdriveN tool. It does feel like there is room for fine tuning but i wouldn't know where to start. Ideally I would like to hit the 2000 sol/s using 115w on gpu-z. Maybe not that low but having a system total power draw of 500w or under for 3 vega 56 would be very nice.
Haha - "sol/s"... someone's been mining ZEC? I have mostly 64s, but my one 56 is running around 1950 H/s at sub-160W ATW. It doesn't seem to be the best asic quality, and is a bit old, so I would expect better from a larger sample set. I also flashed to 64 rom a while ago, which is useless for CNv2 (unless you don't care about power use,) so even lower power use should be expected on a stock rom given that 56 mem runs at a lower voltage, while it's tighter mem timings allow similar results at lower clocks. For 2K H/s (and best efficiency,) I would suggest a voltage setting of around 840mv (or less if you can get away w/ it,) while targeting a 1400 effective core clock. I say effective, because the low voltage will almost certainly cause significant core clock throttling, so you will likely have to set your P7 clock much higher than 1400 to get to that level while under a mining load. For example, I run a 1450 @ 835mv P7 setting, and only get a ~1360 effective clock out of it. On some of my 64s, my clock setting ranges from 1430-1500 (@ 815-850mv) to get to ~1400 effective. As for mem clock, I believe you're looking for a ~950Mhz setting for a 56 - higher is obviously better, if possible. Regarding the PPT, it really is only necessary for two reasons. The first is to lower your P0-5 power state voltages to allow undervolting P6/7 appropriately. This is necessary via PPT, because P0-5 can't be set w/ ODNT/Wattman/etc, but power state voltage settings must be in ascending order. In other words, w/o PPT, the floor for P6/7 voltage is whatever the stock P5 voltage is set to - around 1.1v iirc. With that in mind, I set my P0-P5 @ 800-805mv via PPT, then tune P7 (while miner is running) as appropriate via ODNT. 1450 @ 840mv is prob a good P7 starting point while tuning. The second purpose of PPT is to lock in your final settings once you're done tuning, as it seems the drivers don't seem to always honor the ODNT voltage settings on load, but if those same settings are written to PPT, they 'stick'. This also technically removes the need for loading ODNT profiles at startup, though it doesn't hurt anything to continue to do so. So in short, your steps are: 1. Create a first PPT simply to drop your voltage floor 2. Tune while miner is running to get to an appropriate/stable P7 3. Recreate PPT using final P6/7 values A couple other notes: - a cn_config setting (for trm) of 15+15 seems to be better than 16+14, at least for these lower clocks - CNv2, esp w/ the TRM rocketship, is an unstable beast power wise. Settings which may seem stable at first (even for hours,) may lead to a crash at some point due to power starvation. Even once you think you've found the correct settings, expect to have to dial voltages up (or clocks down) by increments of 5-10 over time to get to a truly stable setup. - You can create PPTs via ODNT, by running in admin mode, then clicking on the GPU icon (top left) and selecting 'PPTable Editor'. - As a starting point, here is my ODNT profile for my 56, though every card, properly tuned, will have different P7 values (also note my mem clock setting is for a 64 bios): [Profile_66] Name=cn8-srvega56-s-r1-2 GPU_P0=852;800 GPU_P1=991;801 GPU_P2=1084;802 GPU_P3=1100;803 GPU_P4=1125;804 GPU_P5=1150;805 GPU_P6=1350;815 GPU_P7=1450;835 Mem_P0=167;800 Mem_P1=500;800 Mem_P2=800;802 Mem_P3=1107;830 Fan_Min=300 Fan_Max=4000 Fan_Target=50 Fan_Acoustic=0 Power_Temp=80 Power_Target=0
EDIT: I always run latest drivers - never had a problem w/ them other than when using stak/derivatives (TRM is an entirely separate codebase, so doesn't suffer from the same issues.) It should be noted that the latest Adrenaline drivers allow setting P0-5 now, which should remove the need for PPT. Sadly, based on my testing it would seem they forgot to extend this capability to the mem power states, which also act as voltage floors to the core.
|
|
|
|
dafuzz_rsa
Newbie
Offline
Activity: 40
Merit: 0
|
|
December 21, 2018, 08:40:08 AM |
|
@pbfarmer Thanks for your insightful post. Helped me shave a few watts off at the wall and increase efficiency a bit
|
|
|
|
heavyarms1912
|
|
December 21, 2018, 07:07:32 PM |
|
EDIT: I always run latest drivers - never had a problem w/ them other than when using stak/derivatives (TRM is an entirely separate codebase, so doesn't suffer from the same issues.) It should be noted that the latest Adrenaline drivers allow setting P0-5 now, which should remove the need for PPT. Sadly, based on my testing it would seem they forgot to extend this capability to the mem power states, which also act as voltage floors to the core.
https://www.dropbox.com/s/ibxcdlrgkm82ccg/OverdriveNTool%200.2.8beta1.7z?dl=1Tried the ODN 2.8 beta?
|
|
|
|
MoneroCrusher
Jr. Member
Offline
Activity: 88
Merit: 1
|
|
December 22, 2018, 11:04:58 AM |
|
I'm suggesting 2 killer features that will make this incredible miner complete: 1. Include in watchdog/logfile the uptime of a crashed GPU (i.e. GPU 4 crashed, uptime: 5 hours, 12 minutes) 2. A simple fan curve in addition to the temperature watchdog (i.e. --temp_max=83,100 -> max temp of 83 while fan must be at 100% and --temp_ramp=75,70 -> when it reaches temp of 75, spin fan to 70% and increase to linearly until max defined temp is reached)
|
|
|
|
nordmann666
Member
Offline
Activity: 363
Merit: 16
|
|
December 22, 2018, 02:57:36 PM |
|
Just a small update on how my small farm is doing, I been running this miner for a week now pretty stable with almost no issues. Not sure how I can check for the percentage of rejected and accepted shares, but all the numbers seem to have evened out. All my miners have been running almost all the same across the board. Maybe a few points of difference between them but they can all hit at least 1900 sol/s for my vega 56's.
A few days ago I read a reddit post of a dude saying he was able to hit 2000 sol/s on his vega 56 with an extremely reduced power draw. He was using the recently released adrenalin drivers but didn't give out much more information. So far I have downloaded the second to latest drivers and have been able to closely replicate the results i had with using the soft powerplay table mod and using overdriveN tool. It does feel like there is room for fine tuning but i wouldn't know where to start. Ideally I would like to hit the 2000 sol/s using 115w on gpu-z. Maybe not that low but having a system total power draw of 500w or under for 3 vega 56 would be very nice.
Haha - "sol/s"... someone's been mining ZEC? I have mostly 64s, but my one 56 is running around 1950 H/s at sub-160W ATW. It doesn't seem to be the best asic quality, and is a bit old, so I would expect better from a larger sample set. I also flashed to 64 rom a while ago, which is useless for CNv2 (unless you don't care about power use,) so even lower power use should be expected on a stock rom given that 56 mem runs at a lower voltage, while it's tighter mem timings allow similar results at lower clocks. For 2K H/s (and best efficiency,) I would suggest a voltage setting of around 840mv (or less if you can get away w/ it,) while targeting a 1400 effective core clock. I say effective, because the low voltage will almost certainly cause significant core clock throttling, so you will likely have to set your P7 clock much higher than 1400 to get to that level while under a mining load. For example, I run a 1450 @ 835mv P7 setting, and only get a ~1360 effective clock out of it. On some of my 64s, my clock setting ranges from 1430-1500 (@ 815-850mv) to get to ~1400 effective. As for mem clock, I believe you're looking for a ~950Mhz setting for a 56 - higher is obviously better, if possible. Regarding the PPT, it really is only necessary for two reasons. The first is to lower your P0-5 power state voltages to allow undervolting P6/7 appropriately. This is necessary via PPT, because P0-5 can't be set w/ ODNT/Wattman/etc, but power state voltage settings must be in ascending order. In other words, w/o PPT, the floor for P6/7 voltage is whatever the stock P5 voltage is set to - around 1.1v iirc. With that in mind, I set my P0-P5 @ 800-805mv via PPT, then tune P7 (while miner is running) as appropriate via ODNT. 1450 @ 840mv is prob a good P7 starting point while tuning. The second purpose of PPT is to lock in your final settings once you're done tuning, as it seems the drivers don't seem to always honor the ODNT voltage settings on load, but if those same settings are written to PPT, they 'stick'. This also technically removes the need for loading ODNT profiles at startup, though it doesn't hurt anything to continue to do so. So in short, your steps are: 1. Create a first PPT simply to drop your voltage floor 2. Tune while miner is running to get to an appropriate/stable P7 3. Recreate PPT using final P6/7 values A couple other notes: - a cn_config setting (for trm) of 15+15 seems to be better than 16+14, at least for these lower clocks - CNv2, esp w/ the TRM rocketship, is an unstable beast power wise. Settings which may seem stable at first (even for hours,) may lead to a crash at some point due to power starvation. Even once you think you've found the correct settings, expect to have to dial voltages up (or clocks down) by increments of 5-10 over time to get to a truly stable setup. - You can create PPTs via ODNT, by running in admin mode, then clicking on the GPU icon (top left) and selecting 'PPTable Editor'. - As a starting point, here is my ODNT profile for my 56, though every card, properly tuned, will have different P7 values (also note my mem clock setting is for a 64 bios): [Profile_66] Name=cn8-srvega56-s-r1-2 GPU_P0=852;800 GPU_P1=991;801 GPU_P2=1084;802 GPU_P3=1100;803 GPU_P4=1125;804 GPU_P5=1150;805 GPU_P6=1350;815 GPU_P7=1450;835 Mem_P0=167;800 Mem_P1=500;800 Mem_P2=800;802 Mem_P3=1107;830 Fan_Min=300 Fan_Max=4000 Fan_Target=50 Fan_Acoustic=0 Power_Temp=80 Power_Target=0
EDIT: I always run latest drivers - never had a problem w/ them other than when using stak/derivatives (TRM is an entirely separate codebase, so doesn't suffer from the same issues.) It should be noted that the latest Adrenaline drivers allow setting P0-5 now, which should remove the need for PPT. Sadly, based on my testing it would seem they forgot to extend this capability to the mem power states, which also act as voltage floors to the core. how you manage to get mem P3 other value than P4 Core? i tried to set mem P3 at 830 but than Core P4 will set also to 830. and are all my cards shit that i cant reach stable 1450Mhz also with 850mV (only 1430 is stable) my main goal is saving energy (to put another Vega into my rig with 1500W PSU)
|
|
|
|
ku4eto
Jr. Member
Offline
Activity: 194
Merit: 4
|
|
December 22, 2018, 09:42:17 PM |
|
I'm suggesting 2 killer features that will make this incredible miner complete: 1. Include in watchdog/logfile the uptime of a crashed GPU (i.e. GPU 4 crashed, uptime: 5 hours, 12 minutes) 2. A simple fan curve in addition to the temperature watchdog (i.e. --temp_max=83,100 -> max temp of 83 while fan must be at 100% and --temp_ramp=75,70 -> when it reaches temp of 75, spin fan to 70% and increase to linearly until max defined temp is reached)
The 18.12 Adrenaline drivers have Fan Curve. But since you are on Linux, that wont help you
|
|
|
|
knittycatkitty
Newbie
Offline
Activity: 19
Merit: 0
|
|
December 23, 2018, 12:03:52 AM |
|
Hi I just came across this new miner!
I'm trying it out right now on my Vega 56 farm.
Switching from cast-xmr on Ubuntu MATE. SupportXMR pool
Will report back with whether I see a improvement or not hashrate, stability and poolside statistics wise
|
|
|
|
kerney666
Member
Offline
Activity: 658
Merit: 86
|
|
December 23, 2018, 04:07:53 AM |
|
I'm suggesting 2 killer features that will make this incredible miner complete: 1. Include in watchdog/logfile the uptime of a crashed GPU (i.e. GPU 4 crashed, uptime: 5 hours, 12 minutes) 2. A simple fan curve in addition to the temperature watchdog (i.e. --temp_max=83,100 -> max temp of 83 while fan must be at 100% and --temp_ramp=75,70 -> when it reaches temp of 75, spin fan to 70% and increase to linearly until max defined temp is reached)
The 18.12 Adrenaline drivers have Fan Curve. But since you are on Linux, that wont help you Yeah, on Linux a fan control mechanism using interpolation between defined points on a fan/temp curve could be a nice addition, not too much work either, it's simple scaling the fan pct using sysfs. For Windows it's getting messier now, I haven't looked into how ADL interacts with the 18.12.2 fan curve at all yet. For the first question, the uptime for all gpus is the same, so it's logged every time the hashrates are logged (every 30 secs or what you set the log interval to). We can add it to the log output made by the watchdog when a gpu is detected dead though, it's trivial really. Personally, I don't see the big value add though .
|
|
|
|
kerney666
Member
Offline
Activity: 658
Merit: 86
|
|
December 23, 2018, 04:10:48 AM |
|
Hi I just came across this new miner!
I'm trying it out right now on my Vega 56 farm.
Switching from cast-xmr on Ubuntu MATE. SupportXMR pool
Will report back with whether I see a improvement or not hashrate, stability and poolside statistics wise
Hi! Sure, let us know if you have any issues. Not to be hard on cast-xmr, but I would expect for you to see a significant boost in both hashrate and efficiency. I assume you're running recent amdgpu-pro drivers (>= 18.30) to begin with?
|
|
|
|
knittycatkitty
Newbie
Offline
Activity: 19
Merit: 0
|
|
December 23, 2018, 07:17:53 AM |
|
Hi I just came across this new miner!
I'm trying it out right now on my Vega 56 farm.
Switching from cast-xmr on Ubuntu MATE. SupportXMR pool
Will report back with whether I see a improvement or not hashrate, stability and poolside statistics wise
Hi! Sure, let us know if you have any issues. Not to be hard on cast-xmr, but I would expect for you to see a significant boost in both hashrate and efficiency. I assume you're running recent amdgpu-pro drivers (>= 18.30) to begin with? Yep 18.30 exact!
|
|
|
|
Kwazii
Newbie
Offline
Activity: 3
Merit: 0
|
|
December 23, 2018, 11:23:54 AM |
|
Hi guys, wath's the api's value (if there's one) that indicates the effective pool hashrate?
Thanks
|
|
|
|
kerney666
Member
Offline
Activity: 658
Merit: 86
|
|
December 23, 2018, 05:48:54 PM |
|
Hi guys, wath's the api's value (if there's one) that indicates the effective pool hashrate?
Thanks
It's called "Work Utility" in the sgminer/cgminer API context. You'll see that passed in the total stats and per gpu. The annoying part is that is needs to be scaled with a multiplier for most algos (but not CN), and is also passed /min, not /sec. So, if you're mining CN: grab the Work Utility API value, divide it /60 so convert from mins to secs, and you should have your poolside hashrate.
|
|
|
|
Kwazii
Newbie
Offline
Activity: 3
Merit: 0
|
|
December 24, 2018, 08:19:45 AM |
|
Great! Thx a lot! It's called "Work Utility" in the sgminer/cgminer API context. You'll see that passed in the total stats and per gpu. The annoying part is that is needs to be scaled with a multiplier for most algos (but not CN), and is also passed /min, not /sec. So, if you're mining CN: grab the Work Utility API value, divide it /60 so convert from mins to secs, and you should have your poolside hashrate.
|
|
|
|
VasilyS
|
|
December 24, 2018, 11:51:13 AM |
|
I'm really surprised by the performance of your miner! It gives hashrate 440 h/s while XMRStack on the same overclocking gives only 370 h/s! Thanks a lot!
|
Information site about technology, cryptocurrencies, radio electronics and other interesting things:
|
|
|
livada
Newbie
Offline
Activity: 417
Merit: 0
|
|
December 24, 2018, 01:21:33 PM |
|
Just a small update on how my small farm is doing, I been running this miner for a week now pretty stable with almost no issues. Not sure how I can check for the percentage of rejected and accepted shares, but all the numbers seem to have evened out. All my miners have been running almost all the same across the board. Maybe a few points of difference between them but they can all hit at least 1900 sol/s for my vega 56's.
A few days ago I read a reddit post of a dude saying he was able to hit 2000 sol/s on his vega 56 with an extremely reduced power draw. He was using the recently released adrenalin drivers but didn't give out much more information. So far I have downloaded the second to latest drivers and have been able to closely replicate the results i had with using the soft powerplay table mod and using overdriveN tool. It does feel like there is room for fine tuning but i wouldn't know where to start. Ideally I would like to hit the 2000 sol/s using 115w on gpu-z. Maybe not that low but having a system total power draw of 500w or under for 3 vega 56 would be very nice.
Haha - "sol/s"... someone's been mining ZEC? I have mostly 64s, but my one 56 is running around 1950 H/s at sub-160W ATW. It doesn't seem to be the best asic quality, and is a bit old, so I would expect better from a larger sample set. I also flashed to 64 rom a while ago, which is useless for CNv2 (unless you don't care about power use,) so even lower power use should be expected on a stock rom given that 56 mem runs at a lower voltage, while it's tighter mem timings allow similar results at lower clocks. For 2K H/s (and best efficiency,) I would suggest a voltage setting of around 840mv (or less if you can get away w/ it,) while targeting a 1400 effective core clock. I say effective, because the low voltage will almost certainly cause significant core clock throttling, so you will likely have to set your P7 clock much higher than 1400 to get to that level while under a mining load. For example, I run a 1450 @ 835mv P7 setting, and only get a ~1360 effective clock out of it. On some of my 64s, my clock setting ranges from 1430-1500 (@ 815-850mv) to get to ~1400 effective. As for mem clock, I believe you're looking for a ~950Mhz setting for a 56 - higher is obviously better, if possible. Regarding the PPT, it really is only necessary for two reasons. The first is to lower your P0-5 power state voltages to allow undervolting P6/7 appropriately. This is necessary via PPT, because P0-5 can't be set w/ ODNT/Wattman/etc, but power state voltage settings must be in ascending order. In other words, w/o PPT, the floor for P6/7 voltage is whatever the stock P5 voltage is set to - around 1.1v iirc. With that in mind, I set my P0-P5 @ 800-805mv via PPT, then tune P7 (while miner is running) as appropriate via ODNT. 1450 @ 840mv is prob a good P7 starting point while tuning. The second purpose of PPT is to lock in your final settings once you're done tuning, as it seems the drivers don't seem to always honor the ODNT voltage settings on load, but if those same settings are written to PPT, they 'stick'. This also technically removes the need for loading ODNT profiles at startup, though it doesn't hurt anything to continue to do so. So in short, your steps are: 1. Create a first PPT simply to drop your voltage floor 2. Tune while miner is running to get to an appropriate/stable P7 3. Recreate PPT using final P6/7 values A couple other notes: - a cn_config setting (for trm) of 15+15 seems to be better than 16+14, at least for these lower clocks - CNv2, esp w/ the TRM rocketship, is an unstable beast power wise. Settings which may seem stable at first (even for hours,) may lead to a crash at some point due to power starvation. Even once you think you've found the correct settings, expect to have to dial voltages up (or clocks down) by increments of 5-10 over time to get to a truly stable setup. - You can create PPTs via ODNT, by running in admin mode, then clicking on the GPU icon (top left) and selecting 'PPTable Editor'. - As a starting point, here is my ODNT profile for my 56, though every card, properly tuned, will have different P7 values (also note my mem clock setting is for a 64 bios): [Profile_66] Name=cn8-srvega56-s-r1-2 GPU_P0=852;800 GPU_P1=991;801 GPU_P2=1084;802 GPU_P3=1100;803 GPU_P4=1125;804 GPU_P5=1150;805 GPU_P6=1350;815 GPU_P7=1450;835 Mem_P0=167;800 Mem_P1=500;800 Mem_P2=800;802 Mem_P3=1107;830 Fan_Min=300 Fan_Max=4000 Fan_Target=50 Fan_Acoustic=0 Power_Temp=80 Power_Target=0
EDIT: I always run latest drivers - never had a problem w/ them other than when using stak/derivatives (TRM is an entirely separate codebase, so doesn't suffer from the same issues.) It should be noted that the latest Adrenaline drivers allow setting P0-5 now, which should remove the need for PPT. Sadly, based on my testing it would seem they forgot to extend this capability to the mem power states, which also act as voltage floors to the core. NICE TUTORIAL. i have vega 56 with 64 bios. Are you sure that V8 algo is better to return stock bios? i have referent vega 56 with samsung mem and now i u se 1095/1100 mhz mem
|
|
|
|
wyzdic
Newbie
Offline
Activity: 46
Merit: 0
|
|
December 25, 2018, 12:39:17 AM |
|
Just a small update on how my small farm is doing, I been running this miner for a week now pretty stable with almost no issues. Not sure how I can check for the percentage of rejected and accepted shares, but all the numbers seem to have evened out. All my miners have been running almost all the same across the board. Maybe a few points of difference between them but they can all hit at least 1900 sol/s for my vega 56's.
A few days ago I read a reddit post of a dude saying he was able to hit 2000 sol/s on his vega 56 with an extremely reduced power draw. He was using the recently released adrenalin drivers but didn't give out much more information. So far I have downloaded the second to latest drivers and have been able to closely replicate the results i had with using the soft powerplay table mod and using overdriveN tool. It does feel like there is room for fine tuning but i wouldn't know where to start. Ideally I would like to hit the 2000 sol/s using 115w on gpu-z. Maybe not that low but having a system total power draw of 500w or under for 3 vega 56 would be very nice.
Haha - "sol/s"... someone's been mining ZEC? I have mostly 64s, but my one 56 is running around 1950 H/s at sub-160W ATW. It doesn't seem to be the best asic quality, and is a bit old, so I would expect better from a larger sample set. I also flashed to 64 rom a while ago, which is useless for CNv2 (unless you don't care about power use,) so even lower power use should be expected on a stock rom given that 56 mem runs at a lower voltage, while it's tighter mem timings allow similar results at lower clocks. For 2K H/s (and best efficiency,) I would suggest a voltage setting of around 840mv (or less if you can get away w/ it,) while targeting a 1400 effective core clock. I say effective, because the low voltage will almost certainly cause significant core clock throttling, so you will likely have to set your P7 clock much higher than 1400 to get to that level while under a mining load. For example, I run a 1450 @ 835mv P7 setting, and only get a ~1360 effective clock out of it. On some of my 64s, my clock setting ranges from 1430-1500 (@ 815-850mv) to get to ~1400 effective. As for mem clock, I believe you're looking for a ~950Mhz setting for a 56 - higher is obviously better, if possible. Regarding the PPT, it really is only necessary for two reasons. The first is to lower your P0-5 power state voltages to allow undervolting P6/7 appropriately. This is necessary via PPT, because P0-5 can't be set w/ ODNT/Wattman/etc, but power state voltage settings must be in ascending order. In other words, w/o PPT, the floor for P6/7 voltage is whatever the stock P5 voltage is set to - around 1.1v iirc. With that in mind, I set my P0-P5 @ 800-805mv via PPT, then tune P7 (while miner is running) as appropriate via ODNT. 1450 @ 840mv is prob a good P7 starting point while tuning. The second purpose of PPT is to lock in your final settings once you're done tuning, as it seems the drivers don't seem to always honor the ODNT voltage settings on load, but if those same settings are written to PPT, they 'stick'. This also technically removes the need for loading ODNT profiles at startup, though it doesn't hurt anything to continue to do so. So in short, your steps are: 1. Create a first PPT simply to drop your voltage floor 2. Tune while miner is running to get to an appropriate/stable P7 3. Recreate PPT using final P6/7 values A couple other notes: - a cn_config setting (for trm) of 15+15 seems to be better than 16+14, at least for these lower clocks - CNv2, esp w/ the TRM rocketship, is an unstable beast power wise. Settings which may seem stable at first (even for hours,) may lead to a crash at some point due to power starvation. Even once you think you've found the correct settings, expect to have to dial voltages up (or clocks down) by increments of 5-10 over time to get to a truly stable setup. - You can create PPTs via ODNT, by running in admin mode, then clicking on the GPU icon (top left) and selecting 'PPTable Editor'. - As a starting point, here is my ODNT profile for my 56, though every card, properly tuned, will have different P7 values (also note my mem clock setting is for a 64 bios): [Profile_66] Name=cn8-srvega56-s-r1-2 GPU_P0=852;800 GPU_P1=991;801 GPU_P2=1084;802 GPU_P3=1100;803 GPU_P4=1125;804 GPU_P5=1150;805 GPU_P6=1350;815 GPU_P7=1450;835 Mem_P0=167;800 Mem_P1=500;800 Mem_P2=800;802 Mem_P3=1107;830 Fan_Min=300 Fan_Max=4000 Fan_Target=50 Fan_Acoustic=0 Power_Temp=80 Power_Target=0
EDIT: I always run latest drivers - never had a problem w/ them other than when using stak/derivatives (TRM is an entirely separate codebase, so doesn't suffer from the same issues.) It should be noted that the latest Adrenaline drivers allow setting P0-5 now, which should remove the need for PPT. Sadly, based on my testing it would seem they forgot to extend this capability to the mem power states, which also act as voltage floors to the core. NICE TUTORIAL. i have vega 56 with 64 bios. Are you sure that V8 algo is better to return stock bios? i have referent vega 56 with samsung mem and now i u se 1095/1100 mhz mem Do not flash to 64, It only increases your power consumption.
|
|
|
|
MoneroCrusher
Jr. Member
Offline
Activity: 88
Merit: 1
|
|
December 25, 2018, 02:01:43 AM Last edit: December 25, 2018, 02:30:19 AM by MoneroCrusher |
|
I'm suggesting 2 killer features that will make this incredible miner complete: 1. Include in watchdog/logfile the uptime of a crashed GPU (i.e. GPU 4 crashed, uptime: 5 hours, 12 minutes) 2. A simple fan curve in addition to the temperature watchdog (i.e. --temp_max=83,100 -> max temp of 83 while fan must be at 100% and --temp_ramp=75,70 -> when it reaches temp of 75, spin fan to 70% and increase to linearly until max defined temp is reached)
The 18.12 Adrenaline drivers have Fan Curve. But since you are on Linux, that wont help you Yeah, on Linux a fan control mechanism using interpolation between defined points on a fan/temp curve could be a nice addition, not too much work either, it's simple scaling the fan pct using sysfs. For Windows it's getting messier now, I haven't looked into how ADL interacts with the 18.12.2 fan curve at all yet. For the first question, the uptime for all gpus is the same, so it's logged every time the hashrates are logged (every 30 secs or what you set the log interval to). We can add it to the log output made by the watchdog when a gpu is detected dead though, it's trivial really. Personally, I don't see the big value add though . So can we expect the fan curve and GPU uptime log ? First one is very helpful for obvious reasons, second one is useful for bigger farms that have to tune many single GPUs. Another suggestion: Split the hw error. i.e. instead of hw: x, you have hw: mem: x, core: x so we can see if an error occured on memory or if it was core related (mostly too low voltage). Would be enormously helpful for debugging/optimizing a larger farm.
|
|
|
|
gerixusyu
Newbie
Offline
Activity: 24
Merit: 0
|
|
December 25, 2018, 08:58:46 PM |
|
Hello everyone,can somebody help me with RX580 8gb Saphire NITRO+, I only get 930mh/s on XMR my clock is 1200 gpu 950mV and 2100 memmory 900mV autoconfig is L8+8
|
|
|
|
Mashy81
Jr. Member
Offline
Activity: 225
Merit: 1
|
|
December 26, 2018, 01:53:15 AM |
|
Hello everyone,can somebody help me with RX580 8gb Saphire NITRO+, I only get 930mh/s on XMR my clock is 1200 gpu 950mV and 2100 memmory 900mV autoconfig is L8+8
Try without the L. Also 8+7 or 7+7
|
|
|
|
|