Claymore-Dual-Miner
Claymore-Dual-Miner copied to clipboard
nanopool
CUDA error 11 - cannot write buffer for DAG
C:\Users\ilhami\Desktop\Ethereum>setx GPU_FORCE_64BIT_PTR 0
SUCCESS: Specified value was saved.
C:\Users\ilhami\Desktop\Ethereum>setx GPU_MAX_HEAP_SIZE 100
SUCCESS: Specified value was saved.
C:\Users\ilhami\Desktop\Ethereum>setx GPU_USE_SYNC_OBJECTS 1
SUCCESS: Specified value was saved.
C:\Users\ilhami\Desktop\Ethereum>setx GPU_MAX_ALLOC_PERCENT 100
SUCCESS: Specified value was saved.
C:\Users\ilhami\Desktop\Ethereum>setx GPU_SINGLE_ALLOC_PERCENT 100
SUCCESS: Specified value was saved.
C:\Users\ilhami\Desktop\Ethereum>EthDcrMiner64.exe -epool eu1.nanopool.org:9999 -ewal "I AM PARANOID, I KNOW NOTHING WILL HAPPEN BUT..." -epsw x -mode 1 -ftime 10
╔════════════════════════════════════════════════════════════════╗ ║ Claymore's Dual ETH + DCR/SC/LBC/PASC GPU Miner v10.0 ║ ╚════════════════════════════════════════════════════════════════╝
This pool (eth-eu2.nanopool.org) does not support Ethereum addresses as login (or requires worker name in "Login.Worker" format) and cannot be used for devfee mining, therefore it is not supported. However, you can mine on this pool if you specify "-allpools 1" option, default pools (different from this pool) will be used for devfee. Please read "Readme" file for details.This pool (eth-eu2.nanopool.org) does not support Ethereum addresses as login (or requires worker name in "Login.Worker" format) and cannot be used for devfee mining, therefore it is not supported. However, you can mine on this pool if you specify "-allpools 1" option, default pools (different from this pool) will be used for devfee. Please read "Readme" file for details.Pool eth-eu2.nanopool.org removed from the list This pool (eth-us-east1.nanopool.org) does not support Ethereum addresses as login (or requires worker name in "Login.Worker" format) and cannot be used for devfee mining, therefore it is not supported. However, you can mine on this pool if you specify "-allpools 1" option, default pools (different from this pool) will be used for devfee. Please read "Readme" file for details.This pool (eth-us-east1.nanopool.org) does not support Ethereum addresses as login (or requires worker name in "Login.Worker" format) and cannot be used for devfee mining, therefore it is not supported. However, you can mine on this pool if you specify "-allpools 1" option, default pools (different from this pool) will be used for devfee. Please read "Readme" file for details.Pool eth-us-east1.nanopool.org removed from the list This pool (eth-us-west1.nanopool.org) does not support Ethereum addresses as login (or requires worker name in "Login.Worker" format) and cannot be used for devfee mining, therefore it is not supported. However, you can mine on this pool if you specify "-allpools 1" option, default pools (different from this pool) will be used for devfee. Please read "Readme" file for details.This pool (eth-us-west1.nanopool.org) does not support Ethereum addresses as login (or requires worker name in "Login.Worker" format) and cannot be used for devfee mining, therefore it is not supported. However, you can mine on this pool if you specify "-allpools 1" option, default pools (different from this pool) will be used for devfee. Please read "Readme" file for details.Pool eth-us-west1.nanopool.org removed from the list This pool (eth-asia1.nanopool.org) does not support Ethereum addresses as login (or requires worker name in "Login.Worker" format) and cannot be used for devfee mining, therefore it is not supported. However, you can mine on this pool if you specify "-allpools 1" option, default pools (different from this pool) will be used for devfee. Please read "Readme" file for details.This pool (eth-asia1.nanopool.org) does not support Ethereum addresses as login (or requires worker name in "Login.Worker" format) and cannot be used for devfee mining, therefore it is not supported. However, you can mine on this pool if you specify "-allpools 1" option, default pools (different from this pool) will be used for devfee. Please read "Readme" file for details.Pool eth-asia1.nanopool.org removed from the list ETH: 1 pool is specified Main Ethereum pool is eu1.nanopool.org:9999 AMD OpenCL platform not found Be careful with overclocking, use default clocks for first tests Press "s" for current statistics, "0".."9" to turn on/off cards, "r" to reload pools, "e" or "d" to select current pool CUDA initializing...
NVIDIA Cards available: 1 CUDA Driver Version/Runtime Version: 9.1/8.0 GPU #0: GeForce GTX 1050, 2048 MB available, 5 compute units, capability: 6.1
Total cards: 1 ETH: Stratum - connecting to 'eu1.nanopool.org' <198.251.88.37> port 9999 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: Stratum - Connected (eu1.nanopool.org:9999) ETH: Authorized Setting DAG epoch #166... Setting DAG epoch #166 for GPU0 Create GPU buffer for GPU0 CUDA error - cannot allocate big buffer for DAG. Check readme.txt for possible solutions. GPU 0 failed Setting DAG epoch #166 for GPU0 GPU 0, CUDA error 11 - cannot write buffer for DAG GPU 0 failed
is the log. Can you guys help me?
Rig: GPU: GTX 1050 CPU: Core i3-7100 Motherboard: Gigabyte H110M-S2V HDD: WD Caviar Blue 1 TB RAM: GSkill RipjawsV 8GB(1x8)
I don't think this is the issue. I have a 1060 with 3GB and I am also getting this error. This just started happening, I was able to mine without an issue a few days ago.
I have exactly the same problem, it started suddenly on February 1. My cards are 1060 3gb just. What I noticed is when ETH is mining , this problem doesnt occur. What so little memory 3 gb allready for the DAG file for ETC?
@ilham123836 @pioniere @rafi213 @m5studio make sure u have a pagefile size 16000-20000.
Still not working.... Set page file to 16GB, still getting the error. Odd that this is suddenly occurring after months of perfect operation, and the DAG file size is < 3GB.
same issue, after a mounts of perfect work rig just refuse to mine i've checked every GPU in rig, see no troubles
Other miners works same way with very low hashrate
Same here with Zotac 1060 3GB card. Since last epoch change in ETC it refuses to work with "cannot allocate big buffer for DAG" error. I tried different pagefile sizes, no fix. Any chance to fix this with a command line parameter? 3GB seems to be an the edge by now...
I had the same issue going from Version 10 to 11. I change 2 or 3 options in the batch file and I think the LIDAG was the important one. It makes the DAG creation go slower....I think. I entered into the batch file: "-lidag 1" (without quotes) Please post if it works for you.
Thanks for the advice, rdavis852, but still no go for ETC with -lidag1 and 0. I tried ETH now and it loads epoch 169 just fine, I think ETC just got too big for 3GB memory :-(
Also got the same issue, tried lidag and gser 2 & 5, dident have this issue before could write dag for 8 gpus same time, but after i´ve connected 3 more gpus it wont work, but it goes further with gser 5 but still not good enought, any solution?
I did some looking around and it seems this issue is caused by Windows 10 not allocating all VRAM to CUDA apps. Many posts say it reserves about 20% memory even if the gpu is a secondary (not used for output).
3GB - 0.6 (20%) = 2.4GB.
Which is about the size of the current ETC DAG 186 (2.45GB) while ETH is still at DAG 176 (2.37GB)
If my analysis is right then ETH mining itself will give the same error when it's DAG size start using more than 2.4GB
Unfortunately, the issue is Windows based and Microsoft knew about it for almost 3 years now and still there is no solution.
I guess people should try Win 7 or 8.1 and see if the issue remains...
Same Problem here also have 1060 3gb Cards. @albalirus your solution worked for me
can also confirm that the '-eres 0'-setting fixed the issue for my 1060 3gb-cards
Can someone post their bat file to show how -eres 0 worked? My 1060 3g rig just started doing the "CUDA error - cannot allocate big buffer for DAG" today and I've tried everything. -eres 0 - 16g vram and nadda
-eres 0 worked up until today's DAG (178)
If you want to continue mining ETH with 1060 3GB, you need to downgrade to Win 7/8.1 or Linux
Set your driver back to V.388.71. It does resolve the problem today. See below for the Nvidia's link for 388.71 driver - http://www.nvidia.com/content/DriverDownload-March2009/confirmation.php?url=%2FWindows%2F388.71%2F388.71-desktop-win10-64bit-international-whql.exe&lang=tw&type=TITAN
@Muaadib same here with zotac 1060 3GB -eres 0 worked until today
@netkingtw will try drivers and report, i got 390.65 atm
Zotac gtx 1060 3 gb. Same error guys -CUDA error 11-cannot write buffer for DAG gpu failed. I m using Windows 10. Somebody please help.
I confirm that nvidia driver 388.71 resolves the issue. Thank you @netkingtw. However, out of eleven nvidia cards, only nine are detected with 388 driver. That's not an issue with 390.xx or 391.xx tho.
I have tried all the things mentioned here until now, just installed the 388.71 drivers as well (and checked if it was applied to everything). Also eres 0, lidag combos, starting two instances of miner to force rebuild dag. Nothing works. Have 10 3gb 1060-s and one 6gb 1060 in the rig.
Any help would be great
@Aljook @remon-georgy
Driver 388.71 works for most miners that had the same cuda error for 1060 3GB GPU yesterday but does not work for a very small part of miners. Taiwan mining pool forum released a solutions to their miners as below:
Step 1, Add the following 5 line commands in your Claymore's miner's .bat file（ -eres 0 should still be added as it was before yesterday's problem occured)：
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
Step 2， Enlarge your rig's virtual memory to at least 32GB.
Step 3, If driver 388.71 does not work for you, use the Windows 10's default Nvidia driver or driver version prior to 388.71 such as 388.13 and make sure that your auto Windows update is disabled .
If all above-mentioned steps still do not resolve your problem, use Claymore 10.6 and try to decrease both core and memory clock a bit until it works.
If all steps fail, use Ethminer instead because Ethminer does not have this issue.
Claymore quote
"It seems 3GB cards won't be able to mine ETH/ETC in Windows10 soon because this OS reserves about 20% of GPU memory size and there is no way to fix it (at least easily). In next update I will set zero devfee for these cards (same as I did for 2GB) so you can mine ETH forks. If you still want to mine ETH/ETC on 3GB cards you should use Windows7 or Linux."
https://bitcointalk.org/index.php?topic=1433925.msg33471755#msg33471755
Seems my theory is correct