 Show Posts
|
|
Pages: [1] 2 3 »
|
|
solved. another mistake in manual here:
HBM2 Example usage: ./amdmemtool -i 0,3,5 --faw 12 --RFC 208
GDDR5 Example usage: ./amdmemtool -i 1,2,4 --RFC 43 --ras2ras 176
-i should be --i
thanks for cool-tool! hope you share your new on-the-fly editor and hope it will accept CLI parameters.
|
|
|
|
|
I use cmd line to set timings for specific gpu:
-i 0 --ACTRD 4 --ACTWR 3 --BUS_TURN 10 --CCLD 2 --CKE 13 --CKE_PULSE 13 --CKSRE 2 --CKSRX 2 --CL 7 --CRCRL 1 --CRCWL 16 --FAW 0 --NOPR 0 --NOPW 0 --PA2RDATA 0 --PA2WDATA 0 --R2R 5 --R2W 15 --RAS 0 --RAS2RAS 28 --RASMACTRD 9 --RASMACTWR 10 --RC 18 --RCDR 5 --RCDRA 5 --RCDW 4 --RCDWA 4 --RFC 39 --RP 10 --RP_RDA 9 --RP_WRA 21 --RRD 1 --SEQ_IDLE 7 --T32AW 2 --TRP 6 --W2R 11 --WDATATR 0 --WRPLUSRP 15
but it returns that timings applied to both gpus - gpu0 and gpu1
Successfully applied new ACTRD, ACTWR, BUS_TURN, CCLD, CKE, CKE_PULSE, CKSRE, CKSRX, CL, CRCRL, CRCWL, FAW, NOPR, NOPW, PA2RDATA, P2WDATA, R2R, R2W, PA2RDATA, RAS2RAS, RASMACTRD, RASMACTWR, RC, RCDR, RCDRA, RCDW, RCDWA, RFC, RP, RP_RDA, RP_WRA, RRD, SEQ_IDLE, T32AW, TRP, W2R, WDATATR, WRPLUSRP settings to GPU 0.
Successfully applied new ACTRD, ACTWR, BUS_TURN, CCLD, CKE, CKE_PULSE, CKSRE, CKSRX, CL, CRCRL, CRCWL, FAW, NOPR, NOPW, PA2RDATA, P2WDATA, R2R, R2W, PA2RDATA, RAS2RAS, RASMACTRD, RASMACTWR, RC, RCDR, RCDRA, RCDW, RCDWA, RFC, RP, RP_RDA, RP_WRA, RRD, SEQ_IDLE, T32AW, TRP, W2R, WDATATR, WRPLUSRP settings to GPU 1.
Is it impossible to apply different timings for each GPU?
|
|
|
|
amdmemtweak --current rasmac twramdmemtweak manual --rasmac wtrmistake?  |
|
|
|
|
In XL version how to import XML profile with CLI? I didnt found any cmd line options (/? or --help).
|
|
|
|
You can also use environment variables in "epools.txt" and "config.txt" files. For example, define "WORKER" environment variable and use it as "%WORKER%" in config.txt or in epools.txt. On Ubuntu (switching because of 1050ti problem) I have environment variable $HOSTNAME: > echo $HOSTNAME Farm9 As said in readme, I specify it in config.txt %HOSTNAME% as my worker but it do not work. |
|
|
|
I've solved the problem with 1050Ti by installing 388.71 driver version
thank you very much, bro! will try it now UPD. It works! But how long? |
|
|
|
Today i waked up to my nvidia rig fo 1050ti 4gb telleing me CUDA error 11 can allocate buffer. Same problem. woke up, 2 farms with 1050ti 4gb are down. claymore 15.0 win10 32gb virtual memory (was 16) but increase not helped Also me. I have four 1050 Ti cards in my farm and as of this morning the stopped working (after the #294 DAG apparently) with following errors: "CUDA error - cannot allocate big buffer for DAG. check readme.txt for possible solutions." "GPU X failed" "GPU X, CUDA error 11 - cannot write buffer for DAG" Reinstalling the drive did not helped. Win10 64bit. 4GB RAM. 32GB Virtual Ram. Claymore v15.0 Please advice, Thank you. same problem here, gt1050ti 4gb. While Claymore keep silence. Have you tried phoenix miner? Well I tried. The same errror, wil not enough memory. Seems like 1050ti 4gb cant use more than 3.3 gb RAM. Latest driver win10 |
|
|
|
Today i waked up to my nvidia rig fo 1050ti 4gb telleing me CUDA error 11 can allocate buffer. Same problem. woke up, 2 farms with 1050ti 4gb are down. claymore 15.0 win10 32gb virtual memory (was 16) but increase not helped NVIDIA Cards available: 1
CUDA Driver Version/Runtime Version: 9.2/8.0
GPU #1: GeForce GTX 1050 Ti, 4096 MB available, 6 compute units, capability: 6.1 (pci bus 1:0:0)
Total cards: 1
You can use "+" and "-" keys to achieve best ETH speed, see "FINE TUNING" section in Readme for details.
GPU #1 strap "2" is applied successfully
All AMD cards use Compute Mode already
CrossFire is disabled already
ETH: Stratum - connecting to 'us1.ethermine.org' <172.65.218.238> port 4444 (unsecure)
ETH: Stratum - Connected (us1.ethermine.org:4444) (unsecure)
ETHEREUM-ONLY MINING MODE ENABLED (-mode 1)
ETH: eth-proxy stratum mode
Watchdog enabled
Remote management (READ-ONLY MODE) is enabled on port 3333
ETH: Authorized
Setting DAG epoch #294(3.30GB)...
ETH: 10/27/19-14:04:01 - New job from us1.ethermine.org:4444
ETH - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0, Time: 00:00
ETH: GPU1 0.000 Mh/s
GPU #1: GeForce GTX 1050 Ti, 4096 MB available, 6 compute units, capability: 6.1 (pci bus 1:0:0)
ETH - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0, Time: 00:00
ETH: GPU1 0.000 Mh/s
Incorrect ETH shares: none
1 minute average ETH total speed: 0.000 Mh/s
Pool switches: ETH - 0
Pool shares accepting time (1 hour stats) - min 0 ms, max 0 ms, average 0 ms
Current ETH share target: 0x0000000112e0be82 (diff: 4000MH), epoch 294(3.30GB)
Current -dcri values: -dcri 30
ETH: 10/27/19-14:04:06 - New job from us1.ethermine.org:4444
ETH - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0, Time: 00:00
ETH: GPU1 0.000 Mh/s
Setting DAG epoch #294 for GPU1
Create GPU buffer for GPU1
CUDA error - cannot allocate big buffer for DAG. Check readme.txt for possible solutions.
ETH: 10/27/19-14:04:09 - New job from us1.ethermine.org:4444
ETH - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0, Time: 00:00
ETH: GPU1 0.000 Mh/s
GPU 1 failed |
|
|
|
|
Miner hangs on strap apply stage if monitor connected to that GPU. If monitor connected to another GPU - everythink fine.
Win10 x64, Sapphire RX470 Hynix 4gb
strap 1 rxboost 1
hanging even with memory clock 1700mhz
|
|
|
|
POLARIS 4GB SAMSUNG straps in Claymore are bad. My custom timings give 29.8, claymores Strap1-4 (RXBOOST=1) maximum 28.8 # STRAP=0 RXBOOST=1 (MY CUSTOM TIMINGS)
# 1950 28.4
# 1970 28.6
# 1990 29
# 2000 29.2
# 2020 29.4
# 2040 29.7
# 2050 29.8 *****
# 2060 29.6
# 2070 crash
# STRAP=6 index out of range
# STRAP=5 index out of range
#STRAP=4#RXBOOST=1#
# 1800 27.7
# 1820 27.9
# 1840 16.8
# 1900 crash
#STRAP=3#RXBOOST=1#
# 1800 27.6
# 1820 27.8
# 1840 28.2
# 1860 28.5
# 1880 28.8
# 1900 15.6
#STRAP=2#RXBOOST=1#
# 1800 27.7
# 1850 28.5
# 1870 28.7
# 1880 28.8
# 1890 26.8
# 1900 17.7
#STRAP=1#RXBOOST=1#
# 1700 26.1
# 1750 26.8
# 1800 27.5
# 1840 28.2
# 1860 28.4
# 1880 28.8
# 1900 16.1
|
|
|
|
|
what better - lower memclock with higher rxboost or higher memclock without rxboost? Asking because rxboost 1 needs lower memclock for stability. So does it worth it?
|
|
|
|
So much tuneups now: -memclock -asm (1,2) -dcri -strape -rxboost  what difference between -rxboost 1 and other values? 1 is equals to what? |
|
|
|
what '-driver install' do? *Driver installed successfully *reboot *GPU-Z show the same driver version  Also I had Samsung Polaris with bat memory ~1860mhz. I flashed stock bios without custom timings, and -strap 1 show only 23mh. Then I flashed-back CUSTOM timings bios and it show 28.5. So maybe miner do not have -straps for polaris_samsung@1860? |
|
|
|
14.0 - 14.2 versions do not detect NVIDIA GPUs on mixed rigs (AMD + NVIDIA)
14.2 No NVIDIA cards in the list, NVML library will not be used. 2:55:53:410 10dc Check and remove old log files...
12:55:53:410 10dc args: -logfile c:\local\FARM6\Miner.log -mode 1 -epool us1.ethermine.org:4444 -ewal 0x********************************.Farm6 -epsw ****** -mpsw ******* -r 1 -ethi 8 -etha 0 -ftime 5 -y 1 -altnum 1 -minspeed 140 -tt 73 -ttdcr 76 -ttli 81 -tstop 85 -fanmin 15 -asm 2
12:55:53:410 10dc
12:55:53:410 10dc ЙНННННННННННННННННННННННННННННННННННННННННННННННННННННННННННННННН»
12:55:53:410 10dc є Claymore's Dual GPU Miner - v14.2 Beta є
12:55:53:410 10dc є ETH + DCR/SIA/LBC/PASC/BLAKE2S/KECCAK є
12:55:53:410 10dc є Supercharged Edition є
12:55:53:410 10dc ИННННННННННННННННННННННННННННННННННННННННННННННННННННННННННННННННј
12:55:53:410 10dc
12:55:53:410 10dc b715
12:55:53:629 10dc ETH: 13 pools are specified
12:55:53:629 10dc Main Ethereum pool is us1.ethermine.org:4444
12:55:54:457 10dc OpenCL platform: NVIDIA CUDA
12:55:54:457 10dc OpenCL platform: AMD Accelerated Parallel Processing
12:55:54:457 10dc OpenCL initializing...
12:55:54:457 10dc AMD Cards available: 3
12:55:54:457 10dc GPU #1: Ellesmere (Radeon (TM) RX 470 Graphics), 4096 MB available, 32 compute units (pci bus 1:0:0)
12:55:54:457 10dc GPU #1 recognized as Radeon RX 470/570
12:55:54:457 10dc GPU #2: Ellesmere (Radeon (TM) RX 470 Graphics), 4096 MB available, 32 compute units (pci bus 2:0:0)
12:55:54:457 10dc GPU #2 recognized as Radeon RX 470/570
12:55:54:457 10dc GPU #3: Ellesmere (Radeon (TM) RX 480 Graphics), 4096 MB available, 36 compute units (pci bus 3:0:0)
12:55:54:457 10dc GPU #3 recognized as Radeon RX 480/580
12:55:54:457 10dc POOL/SOLO version
12:55:54:457 10dc Platform: Windows
12:55:54:535 10dc start building OpenCL program for GPU 1...
12:55:55:035 10dc done
12:55:55:254 10dc start building OpenCL program for GPU 2...
12:55:55:739 10dc done
12:55:55:957 10dc start building OpenCL program for GPU 3...
12:55:56:442 10dc done
12:55:56:707 10dc GPU #1: algorithm ASM 2
12:55:56:754 10dc GPU #2: algorithm ASM 2
12:55:56:817 10dc GPU #3: algorithm ASM 2
12:55:56:832 10dc cudaGetDeviceCount failed (35, CUDA driver version is insufficient for CUDA runtime version), probably no CUDA devices
12:55:56:832 10dc No NVIDIA CUDA GPUs detected.
12:55:56:832 10dc Total cards: 3
12:55:56:832 10dc
Please consider using "-strap" option to increase mining speed, check Readme for details.
12:55:58:864 10dc
You can use "+" and "-" keys to achieve best ETH speed, see "FINE TUNING" section in Readme for details.
12:55:59:879 10dc No NVIDIA cards in the list, NVML library will not be used.
12:55:59:879 10dc All AMD cards use Compute Mode already
12:55:59:879 10dc CrossFire is disabled already
12:56:00:926 10dc SSL: Imported 31 certificates from local storage 12.0 - ok 12:52:51:975 1248 NVIDIA Cards available: 2 12:52:47:663 1248 Check and remove old log files...
12:52:47:663 1248 args: -logfile c:\local\FARM6\Miner.log -mode 1 -epool us1.ethermine.org:4444 -ewal 0x*******.Farm6 -epsw ********** -mpsw ******** -r 1 -ethi 8 -etha 0 -ftime 5 -y 1 -altnum 1 -minspeed 140 -tt 73 -ttdcr 76 -ttli 81 -tstop 85 -fanmin 15 -asm 2
12:52:47:663 1248
12:52:47:663 1248 ЙНННННННННННННННННННННННННННННННННННННННННННННННННННННННННННННННН»
12:52:47:663 1248 є Claymore's Dual GPU Miner - v12.0 є
12:52:47:663 1248 є ETH + DCR/SIA/LBC/PASC/BLAKE2S/KECCAK є
12:52:47:663 1248 ИННННННННННННННННННННННННННННННННННННННННННННННННННННННННННННННННј
12:52:47:663 1248
12:52:47:663 1248 b615
12:52:47:928 1248 ETH: 13 pools are specified
12:52:47:928 1248 Main Ethereum pool is us1.ethermine.org:4444
12:52:48:616 1248 OpenCL platform: NVIDIA CUDA
12:52:48:616 1248 OpenCL platform: AMD Accelerated Parallel Processing
12:52:48:616 1248 OpenCL initializing...
12:52:48:616 1248 AMD Cards available: 3
12:52:48:616 1248 GPU #1: Ellesmere (Radeon (TM) RX 470 Graphics), 4096 MB available, 32 compute units (pci bus 1:0:0)
12:52:48:616 1248 GPU #1 recognized as Radeon RX 470/570
12:52:48:616 1248 GPU #2: Ellesmere (Radeon (TM) RX 470 Graphics), 4096 MB available, 32 compute units (pci bus 2:0:0)
12:52:48:616 1248 GPU #2 recognized as Radeon RX 470/570
12:52:48:616 1248 GPU #3: Ellesmere (Radeon (TM) RX 480 Graphics), 4096 MB available, 36 compute units (pci bus 3:0:0)
12:52:48:616 1248 GPU #3 recognized as Radeon RX 480/580
12:52:48:616 1248 POOL/SOLO version
12:52:48:616 1248 Platform: Windows
12:52:48:756 1248 start building OpenCL program for GPU 1...
12:52:49:241 1248 done
12:52:49:475 1248 start building OpenCL program for GPU 2...
12:52:49:959 1248 done
12:52:50:178 1248 start building OpenCL program for GPU 3...
12:52:50:694 1248 done
12:52:50:959 1248 GPU #1: algorithm ASM 2
12:52:51:022 1248 GPU #2: algorithm ASM 2
12:52:51:084 1248 GPU #3: algorithm ASM 2
12:52:51:975 1248 CUDA initializing...
12:52:51:975 1248 NVIDIA Cards available: 2
12:52:51:975 1248 CUDA Driver Version/Runtime Version: 8.0/8.0
12:52:52:053 1248 GPU #4: GeForce GTX 1070, 8192 MB available, 15 compute units, capability: 6.1 (pci bus 5:0:0)
12:52:52:069 1248 GPU #5: GeForce GTX 1070, 8192 MB available, 15 compute units, capability: 6.1 (pci bus 7:0:0)
12:52:52:069 1248 Total cards: 5
12:52:53:303 1248
You can use "+" and "-" keys to achieve best ETH speed, see "FINE TUNING" section in Readme for details.
12:52:54:381 1248 NVML version: 8.382.53
12:52:54:413 1248 All AMD cards use Compute Mode already
12:52:54:413 1248 CrossFire is disabled already
12:52:56:006 1248 SSL: Imported 31 certificates from local storage |
|
|
|
|
14.0 - 14.2 versions do not detect NVIDIA GPUs on mixed rigs (AMD + NVIDIA)
|
|
|
|
|
> (most tests were performed on 19.4.3).
> miner anyway will install the driver automatically if "-strap" option is used.
|
|
|
|
If I have already flashed the vbios of my 580 and 570 cards, do I need to restore the stock bios before attempting the custom -straps in the command line?
Yes, good question, I'm wondering the same. I think no. This util just replace your bios timings (original or flashed) with eliots. |
|
|
|
until memory clock will go idle (in a few seconds), then enable GPU again in the miner, new clock will be applied after that.
It will be the same as restart miner? I remember that at startup miner try ti find best dcri value, so trick with enabling/disabling GPU with new memclock may be not best way, because optimal DCRI for this card with new memclock may change? Or optimal DCRI may change only if CORE clock have changed? So for new memclock auto_dcri routines is needless |
|
|
|
This magic with straps took me a lot of time and efforts
If you spend so much time, how is the fastest way to found the best strap/mem clock? Is there way to found max mem clock without restarting whole miner? As I remember, when I overclock memory, hashrate go up only after restartminer... so restart miner every time is slow... maybe you found beter way for fast applyong new memclock and see new hashrate without restarting miner? |
|
|
|
Make sure you use "0x" in the beginning of your ETH address. Thank you, Im using system environments and doublecheck echo %ETHWALLET% helped me to understand that it is without 0x. Returned ALLPOOLS: 0 and no warnings now. |
|
|
|
|
