Bitcoin Forum
November 04, 2024, 06:03:05 AM *
News: Latest Bitcoin Core release: 28.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: « 1 ... 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 [320] 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 ... 499 »
  Print  
Author Topic: PhoenixMiner 6.2c: fastest Ethereum/Ethash miner with lowest devfee (Win/Linux)  (Read 784914 times)
ivomm
Legendary
*
Offline Offline

Activity: 1890
Merit: 3094


All good things to those who wait


View Profile
December 04, 2020, 05:17:31 PM
 #6381

Did I already ask for a turbo-kernel, tweak, trick or whatever you want to call it to take advantage of the fact that the memory size of a 8gb card is 3 times the DAG file of the Ethereum Classic dag?

we wont have smart access memory like the new 6800 cards but there must be a possible advantage of that huge free vram.

Same for Ubiq coin or all other ethash coins that use little dag files.

some kind of ramdisk or device-splitting would be nice.....maybe impossible to realize, but I wanted to ask Wink
It is only a matter of time when the devs will take advantage of 6800 architecture. As the author of the notorious video https://www.reddit.com/r/EtherMining/comments/jzp6mp/does_smart_access_memory_unleash_the_6800_non_xt/

pointed out in the comments below:
...
"Additionally it looks like I just stumbled onto a bug which enabled what I have shown in the video. It's deeper than SAM and looks like it might be tied weirdly to Adobe Premiere's OpenCL render engine, where it jacks up something with OpenCL to allow two independent 'device' id's, that allow the miner to page to both as independent GPUs, opening up two separate DAG files in memory."

So without any optimisation this 'bug' allowed two threads 41 Mh each vs one thread 62 Mh with the default settings. The increase of the hash is not linear, but we should expect may be 100 Mh for 3 threads, 120 for 4 threads, etc. Of course, devs will need some time to do it. And I expect some other optimisations, which may give another 10-20% boost. So IMO 6800 is quite a perspective investment, provided that ethereum will be mineable for the next 12 months.
lembzkie
Newbie
*
Offline Offline

Activity: 1
Merit: 0


View Profile
December 04, 2020, 05:54:54 PM
 #6382

2020.12.05:01:46:32.191: GPU1 GPU1: Allocating DAG for epoch #379 (3.93) GB
2020.12.05:01:46:32.191: GPU1 GPU1: Allocating buffers failed with: clCreateBuffer (-61).
2020.12.05:01:46:32.191: wdog Fatal error detected. Restarting.

Can anyone help me please?
Cryptomining2003
Newbie
*
Offline Offline

Activity: 64
Merit: 0


View Profile
December 04, 2020, 06:10:50 PM
 #6383

Did I already ask for a turbo-kernel, tweak, trick or whatever you want to call it to take advantage of the fact that the memory size of a 8gb card is 3 times the DAG file of the Ethereum Classic dag?

we wont have smart access memory like the new 6800 cards but there must be a possible advantage of that huge free vram.

Same for Ubiq coin or all other ethash coins that use little dag files.

some kind of ramdisk or device-splitting would be nice.....maybe impossible to realize, but I wanted to ask Wink
It is only a matter of time when the devs will take advantage of 6800 architecture. As the author of the notorious video https://www.reddit.com/r/EtherMining/comments/jzp6mp/does_smart_access_memory_unleash_the_6800_non_xt/

pointed out in the comments below:
...
"Additionally it looks like I just stumbled onto a bug which enabled what I have shown in the video. It's deeper than SAM and looks like it might be tied weirdly to Adobe Premiere's OpenCL render engine, where it jacks up something with OpenCL to allow two independent 'device' id's, that allow the miner to page to both as independent GPUs, opening up two separate DAG files in memory."

So without any optimisation this 'bug' allowed two threads 41 Mh each vs one thread 62 Mh with the default settings. The increase of the hash is not linear, but we should expect may be 100 Mh for 3 threads, 120 for 4 threads, etc. Of course, devs will need some time to do it. And I expect some other optimisations, which may give another 10-20% boost. So IMO 6800 is quite a perspective investment, provided that ethereum will be mineable for the next 12 months.

good point...thx for that....but I wanted to point at the older series where you have huge free vram too......not just the new 6800 series, the 5000 series e.g. with the decrease of the dag for etc we have this new option. eth dag cant be loaded twice, but etc / ubiq can. I think there was a kernel called turbo for this situation. that needs to be updated for the 5000 series or nvidia gpu´s with 8+ gb vram.
stonetown
Newbie
*
Offline Offline

Activity: 3
Merit: 0


View Profile
December 04, 2020, 06:22:01 PM
 #6384

PhoenixMiner 5.3b on Linux no control for clock and voltage

I have rig 8 x rx580 and trying to move from windows to linux.

I have installed Ubuntu 18.04.5 HWE with kernel 5.6.0 and  amdgpu-pro-20.20-1089974-ubuntu-18.04. (./amdgpu-pro-install -y --opencl=pal,legacy --headless)

It seems that miner can not control clock and voltage

sudo ./PhoenixMiner -tt 59 -gpus 1  -mode 1 -hstats 2 -cclock 1180 -mclock 2150 -cvddc 900 -mvddc 900 -pool us2.ethermine.org:4444 -wal 0x39e987c72d3b35da03ba9aca26a9e9f91c16c04c.Rig001
Phoenix Miner 5.3b Linux/gcc - Release build
--------------------------------------------

No CUDA driver found
OpenCL driver version: 20.20-1089974-ubuntu-18.04
Available GPUs for mining:
GPU1: Radeon RX 580 Series (pcie 1), OpenCL 1.2, 8 GB VRAM, 36 CUs
Eth: Loading pools from epools.txt
Eth: the pool list contains 6 pools (1 from command-line)
Eth: primary pool: us2.ethermine.org:4444
Starting GPU mining
GPU1: fan PWM control mode 2, min 0, max 255
GPU1: set auto fan: 59C target temp (min fan 0, max fan 100)
Eth: Connecting to ethash pool us2.ethermine.org:4444 (proto: EthProxy)

...
Eth speed: 26.297 MH/s, shares: 0/0/0, time: 0:00
Eth: New job #39a2a3b1 from us2.ethermine.org:4444; diff: 4000MH
GPU1: 42C 0% 107W
GPU1: cclock 1340 MHz, cvddc 1131 mV, mclock 2000 MHz
GPUs power: 107.1 W

...
Eth: New job #8993d4fa from us2.ethermine.org:4444; diff: 4000MH
Eth speed: 28.457 MH/s, shares: 1/0/0, time: 0:02
GPU1: 59C 69% 118W
GPU1: cclock 1340 MHz, cvddc 1143 mV, mclock 2000 MHz
GPUs power: 118.2 W


What is wrong? any help/advice would be greatly appreciated

zcyber
Newbie
*
Offline Offline

Activity: 5
Merit: 0


View Profile
December 04, 2020, 06:36:22 PM
 #6385

any news on vega hashrate drops?

Very interested in this too. Seems that straps not working anymore.

My x5Vegas56 works at around 42.5mhs for now on ETH (852mhz800mv 930mhz/800mv) and x3 rx574 around 29.2 mhs on ETC (1150mhz810mv 2225mhz810mv). whatever i do rxboost on or off nothing changes but straps worked for Vegas (PM sure with admin rights). It was 44+mhs on these settings and it was really great because after lots of tuning my 8gpu rig takes only 1150watt for 300mhs.

I use 2 .bat files. one for Vega56 on ETH and second for rx570 on ETC on same system (windows 10) looks stable enough. No leak of hashrate but less ETH. Still more profitable I think.

For all who don`t know how to calculate energy consumption - just add up all the indicators with the watt "W" sign for each card in the HWinfo. I compared results with hardware meter and it shows same numbers.

Have no idea how it possible to squeeze 50+mhs from Vega56 on showed upper settings. I can`t try Linux because one of my vega56 shows only 54CUs and Linux cant recognize it correctly, but on windows seems it is impossible or it will be very unstable and too hot on 200+watts for GPU.

kaiokenx10
Newbie
*
Offline Offline

Activity: 21
Merit: 0


View Profile
December 04, 2020, 07:27:50 PM
Last edit: December 04, 2020, 08:16:31 PM by kaiokenx10
 #6386

any news on vega hashrate drops?

Very interested in this too. Seems that straps not working anymore.

My x5Vegas56 works at around 42.5mhs for now on ETH (852mhz800mv 930mhz/800mv) and x3 rx574 around 29.2 mhs on ETC (1150mhz810mv 2225mhz810mv). whatever i do rxboost on or off nothing changes but straps worked for Vegas (PM sure with admin rights). It was 44+mhs on these settings and it was really great because after lots of tuning my 8gpu rig takes only 1150watt for 300mhs.

I use 2 .bat files. one for Vega56 on ETH and second for rx570 on ETC on same system (windows 10) looks stable enough. No leak of hashrate but less ETH. Still more profitable I think.

For all who don`t know how to calculate energy consumption - just add up all the indicators with the watt "W" sign for each card in the HWinfo. I compared results with hardware meter and it shows same numbers.

Have no idea how it possible to squeeze 50+mhs from Vega56 on showed upper settings. I can`t try Linux because one of my vega56 shows only 54CUs and Linux cant recognize it correctly, but on windows seems it is impossible or it will be very unstable and too hot on 200+watts for GPU.



something is really messed up right now, i just noticed that i can't open hwinfo while my gpus are running, the rig crashes

edit: i switched to simple mining, the hashrate is back where it should be Smiley
abrgeza
Jr. Member
*
Offline Offline

Activity: 117
Merit: 2


View Profile
December 04, 2020, 09:44:49 PM
 #6387

PhoenixMiner 5.3b on Linux no control for clock and voltage

I have rig 8 x rx580 and trying to move from windows to linux.
I have installed Ubuntu 18.04.5 HWE with kernel 5.6.0 and  amdgpu-pro-20.20-1089974-ubuntu-18.04. (./amdgpu-pro-install -y --opencl=pal,legacy --headless)
It seems that miner can not control clock and voltage

What is wrong? any help/advice would be greatly appreciated



Did you read ReleaseNotes?
... Implemented hardware control (setting clocks, voltages, power limit, fan control) for AMD cards under Linux.
    Requires PhoenixMiner running with root permissions. The clocks and voltage control require relatively
    new drivers (tested on 19.30-855429) and adding the following boot parameter to Linux kernel:
      amdgpu.ppfeaturemask=0xffffffff ...

The next question will be how to do that.  Smiley
Geniusvzla
Newbie
*
Offline Offline

Activity: 36
Merit: 0


View Profile
December 05, 2020, 12:23:12 AM
 #6388

Just saw your post, I get it that you were using stock bios cards with Claymore and applying straps and rxboost to increase the hashrate. AFAIK Phoenix miner currently does not support "straps" for RX580 (all polaris) cards, so you'll have to bios mod these cards manually (using 1 click bios mods or any other tools) and then try using the overdriventool settings and then you'll get similar hash rates on RX580s.


I flashed the cards with a new BIOS and I now get 31.2 Mh/s on the RX580.
Thanks for the Steer on not using Straps!


Seems to be mining fine at the moment after I tweaked the clocks a little. One thing I did find though is that if PhoenixMiner hash rate on any card goes to 0Mh/s (due to too much overclocking), when it restarts, it can't seem to do anything with the card. It connects and 'sees' the card but it is ignored from the autotune and restarts at 0MH/s. The only way round I found was to reflash the card BIOS...very strange. Is that normal? Or is there some other way to tell PhoenixMiner to forget it had an issue with the card previously and to start from scratch again?

Very pleased that there's an alternative to Claymore and, it seems more stable (apart from the issue mentioned above).

Greetings, the same happened to me and I had to change the bios of each card that failed me, update the AMD driver to the latest version and now I control the voltage and memory with the Overdriventool program, which by the way is Excellent. I create a .bat file so that when the platforms are rebooted they will keep the settings I put in and I only use MSI Afterburner to control the ventilation.
I get about 33 Mh / s with RX580 nitro 8GB (Micron memory) and do not generate memory errors. So everything is going well so far Smiley
Sorry for my bad English, everything is translation Smiley
stonetown
Newbie
*
Offline Offline

Activity: 3
Merit: 0


View Profile
December 05, 2020, 02:17:23 AM
 #6389

PhoenixMiner 5.3b on Linux no control for clock and voltage

I have rig 8 x rx580 and trying to move from windows to linux.
I have installed Ubuntu 18.04.5 HWE with kernel 5.6.0 and  amdgpu-pro-20.20-1089974-ubuntu-18.04. (./amdgpu-pro-install -y --opencl=pal,legacy --headless)
It seems that miner can not control clock and voltage

What is wrong? any help/advice would be greatly appreciated



Did you read ReleaseNotes?
... Implemented hardware control (setting clocks, voltages, power limit, fan control) for AMD cards under Linux.
    Requires PhoenixMiner running with root permissions. The clocks and voltage control require relatively
    new drivers (tested on 19.30-855429) and adding the following boot parameter to Linux kernel:
      amdgpu.ppfeaturemask=0xffffffff ...

The next question will be how to do that.  Smiley


Yes. implementing amdgpu.ppfeaturemask=0xffffffff helped. Everything works now. Thanks a lot for your help
Cryptomining2003
Newbie
*
Offline Offline

Activity: 64
Merit: 0


View Profile
December 05, 2020, 07:37:36 AM
 #6390

i suggest a mining os like phoenix recommends in his first post.
restless
Legendary
*
Offline Offline

Activity: 1151
Merit: 1001


View Profile
December 05, 2020, 09:56:19 AM
 #6391

2020.12.05:01:46:32.191: GPU1 GPU1: Allocating DAG for epoch #379 (3.93) GB
2020.12.05:01:46:32.191: GPU1 GPU1: Allocating buffers failed with: clCreateBuffer (-61).
2020.12.05:01:46:32.191: wdog Fatal error detected. Restarting.

Can anyone help me please?
Its like that since 2nd December... many people write here, but no one gave solution
Latest Hive OS with me, rig of 4 radeons - 8gb main one, 4gb the rest.
bategojko74
Member
**
Offline Offline

Activity: 220
Merit: 12


View Profile
December 05, 2020, 11:44:40 AM
 #6392

2020.12.05:01:46:32.191: GPU1 GPU1: Allocating DAG for epoch #379 (3.93) GB
2020.12.05:01:46:32.191: GPU1 GPU1: Allocating buffers failed with: clCreateBuffer (-61).
2020.12.05:01:46:32.191: wdog Fatal error detected. Restarting.

Can anyone help me please?
Its like that since 2nd December... many people write here, but no one gave solution
Latest Hive OS with me, rig of 4 radeons - 8gb main one, 4gb the rest.
Update your driver.
Switch 4GB cards to ETC.
Bird333
Newbie
*
Offline Offline

Activity: 4
Merit: 0


View Profile
December 05, 2020, 12:18:58 PM
 #6393

Hello ,can Any Help Smiley
I switched Clay to Phoenixminer 5.3, I have Saphire and XFX rx580 8gb, adrenaline 19.4.3 driver, Win 10 ,  on Clay everything worked normally for me, now 3xGraphic will not switch 28.8 Here is the command line in file.bat
setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
timeout 15
PhoenixMiner.exe -pool ssl: //eu1.ethermine.org: 5555 -pool2 ssl: //us1.ethermine.org: 5555 -wal ******************* ******** -proto 3 -retrydelay 5 -logsmaxsize 0 -acm -cvddc 850 -cclock 1150,1160,1160,1160,1160,1150,1160,1150 -mvddc 850 -mclock 2080,2150,2150, 2150,2150,2150,2150,2150 -tt 55 -tstart 50 -tstop 75 -fanmin 30 -fanmax 80 -hstats 2 -wdog 0 -rxboost 1 -log 0
pause

https://imgur.com/a/fLY3A3k
mraychev
Newbie
*
Offline Offline

Activity: 1
Merit: 0


View Profile
December 05, 2020, 01:06:53 PM
 #6394

Hello,

I switched to phoneix 5.3, but I notice an increased energy consultation.

I have platforms 2x6 shapshire rx 470 8g - 350 mh / s, 1900watts! at cleymore 15 - 1500watts
1x 5 rx 580 powercolor - 150 mh / s / 920watts - claymore 15 800watts

please advice and settings.

SAPPHIRE NITRO+ Radeon RX 470 8GB

2020.12.04: 03: 17: 30.750: main No CUDA driver found
2020.12.04: 03: 17: 33.584: main OpenCL driver version: 17.31.0
2020.12.04: 03: 17: 33.624: main Available GPUs for mining:
2020.12.04: 03: 17: 33.624: main GPU1: Radeon (TM) RX 470 Graphics (pcie 1), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU2: Radeon (TM) RX 470 Graphics (pcie 2), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU3: Radeon (TM) RX 470 Graphics (pcie 3), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU4: Radeon (TM) RX 470 Graphics (pcie 4), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU5: Radeon (TM) RX 470 Graphics (pcie 5), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU6: Radeon (TM) RX 470 Graphics (pcie 6), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.625: main ADL library initialized

RX 580 8gb powercolor
2020.12.04:13:18:02.423: main Phoenix Miner 5.3b Windows/msvc - Release build
2020.12.04:13:18:02.423: main No CUDA driver found
2020.12.04:13:18:03.769: main OpenCL driver version: 17.31.0
2020.12.04:13:18:03.790: main Available GPUs for mining:
2020.12.04:13:18:03.790: main GPU1: Radeon RX 580 Series (pcie 1), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.790: main GPU2: Radeon RX 580 Series (pcie 3), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.790: main GPU3: Radeon RX 580 Series (pcie 5), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.790: main GPU4: Radeon RX 580 Series (pcie 7), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.790: main GPU5: Radeon RX 580 Series (pcie Cool, OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.791: main ADL library initialized

Do I need to change the OpenCL driver or driver for card:

bat file:

@echo off
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_HEAP_SIZE 100
setx GPU_IMAGE_DMA 1
setx GPU_ENABLE_LARGE_ALLOCATION 1
setx GPU_FORCE_64BIT_PTR 0

PhoenixMiner.exe -pool eth-eu1.nanopool.org:9999 -wal
-amd
-acm
-mode 1
-hstats 1
-cclock 1150
-mclock 2040
-cvddc 900
-mvddc 850
-log 0
-tstop 85
-tstart 55
-tt 65
-fanmin 45
-fanmax 88
-cdm 0
-fcm 0
-gpus 1,2,3,4,5,6,7
-clKernel 1
-gser 5
-rvram -1
-wdog 1
-rmode 0
-powlim -50
-mt 2
I may have put something I shouldn't
They were not used for a very long time and worked only on Claymore. Please tell me which one to remove or put on to reduce electricity consumption

Settings of afterburner:
core voltage - /-98/
power limit - /-16/
temp. limit - /78/
cclock - 1150
mmclock - 2100/for rx580/; for rx470 - 2050





1. download and install and drivers 19.9
2. open and settings go to every gpu and witch from graphics to compute mode
this should work... I guess you are using the 17 blockchain beta driver now
GoblinCBR
Newbie
*
Offline Offline

Activity: 5
Merit: 0


View Profile
December 05, 2020, 01:43:31 PM
 #6395

is there any command line that can restart phoenix miner after 24 hours ? I used to do that in claymore but I cannot find the line in the readme file of phoenix . thanks
Guigs321
Newbie
*
Offline Offline

Activity: 41
Merit: 0


View Profile
December 05, 2020, 02:54:53 PM
 #6396

is there any command line that can restart phoenix miner after 24 hours ? I used to do that in claymore but I cannot find the line in the readme file of phoenix . thanks

-timeout 1440 (for restart the miner each 24h) -rmode1 (to restart with the same settings)
zlatarovv
Newbie
*
Offline Offline

Activity: 2
Merit: 0


View Profile
December 05, 2020, 08:08:21 PM
 #6397

Hello,

I switched to phoneix 5.3, but I notice an increased energy consultation.

I have platforms 2x6 shapshire rx 470 8g - 350 mh / s, 1900watts! at cleymore 15 - 1500watts
1x 5 rx 580 powercolor - 150 mh / s / 920watts - claymore 15 800watts

please advice and settings.

SAPPHIRE NITRO+ Radeon RX 470 8GB

2020.12.04: 03: 17: 30.750: main No CUDA driver found
2020.12.04: 03: 17: 33.584: main OpenCL driver version: 17.31.0
2020.12.04: 03: 17: 33.624: main Available GPUs for mining:
2020.12.04: 03: 17: 33.624: main GPU1: Radeon (TM) RX 470 Graphics (pcie 1), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU2: Radeon (TM) RX 470 Graphics (pcie 2), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU3: Radeon (TM) RX 470 Graphics (pcie 3), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU4: Radeon (TM) RX 470 Graphics (pcie 4), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU5: Radeon (TM) RX 470 Graphics (pcie 5), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU6: Radeon (TM) RX 470 Graphics (pcie 6), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.625: main ADL library initialized

RX 580 8gb powercolor
2020.12.04:13:18:02.423: main Phoenix Miner 5.3b Windows/msvc - Release build
2020.12.04:13:18:02.423: main No CUDA driver found
2020.12.04:13:18:03.769: main OpenCL driver version: 17.31.0
2020.12.04:13:18:03.790: main Available GPUs for mining:
2020.12.04:13:18:03.790: main GPU1: Radeon RX 580 Series (pcie 1), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.790: main GPU2: Radeon RX 580 Series (pcie 3), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.790: main GPU3: Radeon RX 580 Series (pcie 5), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.790: main GPU4: Radeon RX 580 Series (pcie 7), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.790: main GPU5: Radeon RX 580 Series (pcie Cool, OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.791: main ADL library initialized

Do I need to change the OpenCL driver or driver for card:

bat file:

@echo off
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_HEAP_SIZE 100
setx GPU_IMAGE_DMA 1
setx GPU_ENABLE_LARGE_ALLOCATION 1
setx GPU_FORCE_64BIT_PTR 0

PhoenixMiner.exe -pool eth-eu1.nanopool.org:9999 -wal
-amd
-acm
-mode 1
-hstats 1
-cclock 1150
-mclock 2040
-cvddc 900
-mvddc 850
-log 0
-tstop 85
-tstart 55
-tt 65
-fanmin 45
-fanmax 88
-cdm 0
-fcm 0
-gpus 1,2,3,4,5,6,7
-clKernel 1
-gser 5
-rvram -1
-wdog 1
-rmode 0
-powlim -50
-mt 2
I may have put something I shouldn't
They were not used for a very long time and worked only on Claymore. Please tell me which one to remove or put on to reduce electricity consumption

Settings of afterburner:
core voltage - /-98/
power limit - /-16/
temp. limit - /78/
cclock - 1150
mmclock - 2100/for rx580/; for rx470 - 2050





1. download and install and drivers 19.9
2. open and settings go to every gpu and witch from graphics to compute mode
this should work... I guess you are using the 17 blockchain beta driver now

Dear all,
I changed the drivers, but there is no change in power consumption?
If you have other suggestions, the cards consume over 110 watts
GoblinCBR
Newbie
*
Offline Offline

Activity: 5
Merit: 0


View Profile
December 05, 2020, 08:15:38 PM
 #6398

is there any command line that can restart phoenix miner after 24 hours ? I used to do that in claymore but I cannot find the line in the readme file of phoenix . thanks

-timeout 1440 (for restart the miner each 24h) -rmode1 (to restart with the same settings)

Thanks a lot men
miner29
Full Member
***
Offline Offline

Activity: 1281
Merit: 141


View Profile
December 05, 2020, 08:21:14 PM
 #6399

Quick reminder...only PhoenixMiner is where you get updates for PhoenixMine.

Scammers try to link new versions they claim are ne releases...It is malware.

Do not fall for scammers....only download from posts from PM himself.  Or any other miner dev.  Never from some rando user.
DrX
Member
**
Offline Offline

Activity: 233
Merit: 20


View Profile
December 05, 2020, 09:34:49 PM
 #6400

Code:
[quote author=zlatarovv link=topic=2647654.msg55760704#msg55760704 date=1607198901]
[quote author=mraychev link=topic=2647654.msg55757515#msg55757515 date=1607173613]
[quote author=zlatarovv link=topic=2647654.msg55749355#msg55749355 date=1607082236]
Hello,

I switched to phoneix 5.3, but I notice an increased energy consultation.

I have platforms 2x6 shapshire rx 470 8g - 350 mh / s, 1900watts! at cleymore 15 - 1500watts
1x 5 rx 580 powercolor - 150 mh / s / 920watts - claymore 15 800watts

please advice and settings.

SAPPHIRE NITRO+ Radeon RX 470 8GB

2020.12.04: 03: 17: 30.750: main No CUDA driver found
2020.12.04: 03: 17: 33.584: main OpenCL driver version: 17.31.0
2020.12.04: 03: 17: 33.624: main Available GPUs for mining:
2020.12.04: 03: 17: 33.624: main GPU1: Radeon (TM) RX 470 Graphics (pcie 1), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU2: Radeon (TM) RX 470 Graphics (pcie 2), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU3: Radeon (TM) RX 470 Graphics (pcie 3), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU4: Radeon (TM) RX 470 Graphics (pcie 4), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU5: Radeon (TM) RX 470 Graphics (pcie 5), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.624: main GPU6: Radeon (TM) RX 470 Graphics (pcie 6), OpenCL 2.0, 8 GB VRAM, 32 CUs
2020.12.04: 03: 17: 33.625: main ADL library initialized

RX 580 8gb powercolor
2020.12.04:13:18:02.423: main Phoenix Miner 5.3b Windows/msvc - Release build
2020.12.04:13:18:02.423: main No CUDA driver found
2020.12.04:13:18:03.769: main OpenCL driver version: 17.31.0
2020.12.04:13:18:03.790: main Available GPUs for mining:
2020.12.04:13:18:03.790: main GPU1: Radeon RX 580 Series (pcie 1), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.790: main GPU2: Radeon RX 580 Series (pcie 3), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.790: main GPU3: Radeon RX 580 Series (pcie 5), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.790: main GPU4: Radeon RX 580 Series (pcie 7), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.790: main GPU5: Radeon RX 580 Series (pcie 8), OpenCL 2.0, 8 GB VRAM, 36 CUs
2020.12.04:13:18:03.791: main ADL library initialized

Do I need to change the OpenCL driver or driver for card:

bat file:

@echo off
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_HEAP_SIZE 100
setx GPU_IMAGE_DMA 1
setx GPU_ENABLE_LARGE_ALLOCATION 1
setx GPU_FORCE_64BIT_PTR 0

PhoenixMiner.exe -pool eth-eu1.nanopool.org:9999 -wal
-amd
-acm
-mode 1
-hstats 1
-cclock 1150
-mclock 2040
-cvddc 900
-mvddc 850
-log 0
-tstop 85
-tstart 55
-tt 65
-fanmin 45
-fanmax 88
-cdm 0
-fcm 0
-gpus 1,2,3,4,5,6,7
-clKernel 1
-gser 5
-rvram -1
-wdog 1
-rmode 0
-powlim -50
-mt 2
I may have put something I shouldn't
They were not used for a very long time and worked only on Claymore. Please tell me which one to remove or put on to reduce electricity consumption

Settings of afterburner:
core voltage - /-98/
power limit - /-16/
temp. limit - /78/
cclock - 1150
mmclock - 2100/for rx580/; for rx470 - 2050




[/quote]

1. download and install and drivers 19.9
2. open and settings go to every gpu and witch from graphics to compute mode
this should work... I guess you are using the 17 blockchain beta driver now
[/quote]

Dear all,
I changed the drivers, but there is no change in power consumption?
If you have other suggestions, the cards consume over 110 watts
[/quote]

You set same settings in Afterburner and in PM? You don't need AB if you use PM to set the settings. What wattage does PM report?
Pages: « 1 ... 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 [320] 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 ... 499 »
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!