Author

Topic: ethminer refuses to mine ETH with "out of memory on CUDA device" error (Read 146 times)

legendary
Activity: 1820
Merit: 1121
Thanks for your advice, guys. I downloaded and am currently using something called lolminer, like the first person suggested. It works, but the hash rate is very bad. Plus, I can imagine that they are taking some of it for the developer. since it's not a free miner like ethminer used to be.

Nevertheless, I guess it doesn't matter as ETH mining will cease to exist in a couple of days, right?
Russian miners board have information about this, I will translate:
The hash rate drops on WIN 10 by 6 GB NVIDIA - install Gminer 3.05, driver version 512.16
https://bitcointalksearch.org/topic/m.60663759
But since there are less than 18 hours left before the end of ethereum mining, this will not help you much. On older video cards, ethminer works fine.
legendary
Activity: 2366
Merit: 1408
Thanks for your advice, guys. I downloaded and am currently using something called lolminer, like the first person suggested. It works, but the hash rate is very bad. Plus, I can imagine that they are taking some of it for the developer. since it's not a free miner like ethminer used to be.

Nevertheless, I guess it doesn't matter as ETH mining will cease to exist in a couple of days, right?

I think the fee of lolminer to mine ETH is only 0.7%, it's not high

Hou much your cards are making?
As I said, you can point to ETC just to check if the dag size is reducing your hashrate
jr. member
Activity: 49
Merit: 1
Thanks for your advice, guys. I downloaded and am currently using something called lolminer, like the first person suggested. It works, but the hash rate is very bad. Plus, I can imagine that they are taking some of it for the developer. since it's not a free miner like ethminer used to be.

Nevertheless, I guess it doesn't matter as ETH mining will cease to exist in a couple of days, right?
legendary
Activity: 2534
Merit: 6080
Self-proclaimed Genius
So a few days back I got a notification from the mining pool that it hasn't received any good hashes. I checked my mining rig, rebooted it, and it was showing the following errors upon the start of the ethminer:
Code:
ethminer 0.19.0-alpha.0
...
cu 06:27:11 cuda-1   Unexpected error CUDA error in func dev::eth::CUDAMiner::initEpoch_internal at line 123 out of memory on CUDA device 03:00.0

Any idea what is going on there?
That's quite an old miner, AFAIK ethminer 0.19.0 is from 2019.
That said, most of these Ethereum miners can only work up to a certain epoch, there's a high chance that that specific version can only work up to the 515th.

For example (can be related), the old Claymore ETH miner has this in the official thread:
NOTE: please upgrade to v15.0 as soon as possible, old versions support up to #299 epoch only, they will fail then!
v15.0 supports up to #384 epoch (4GB DAG size).

That said, you may need to switch to an updated miner.
legendary
Activity: 2366
Merit: 1408
Can you try a new miner?
I recommend T-Rex: https://bitcointalksearch.org/topic/t-rex-0268-ethw-etc-dual-mining-gpu-miner-4432704
or Gminer: https://bitcointalksearch.org/topic/gminer-v324-ergokaspazilkawpowequihashcortex-5034735
Nbminer is also good

I suspect you can solve this issue with an updated miner

To check if the problem is with ETH dag size, you can point your ethminer to mine ETC, if works, it's the dag size, since dag of ETC is smaller.
member
Activity: 759
Merit: 15
Unfortunately the dag to mine eth has increased it is not more than about 4gb but it is more than 5gb your gpu have 6gb of memory but a part of it is in common use with the system so it is not enough you can try lolminer with the zombie mode but still the hash rate will be very low
legendary
Activity: 3318
Merit: 1247
Bitcoin Casino Est. 2013
Nothing going on except you are using not the best of the miners,try with lolminer or with Nbminer and if you get the same error,try updating your driver to the latest version.

However don't expect much from this card it was doing 24-25 Mhsh in the beginnings but now with the Eth dag size over 5 Gb this will do barely 19-20 from what I know.Anyway try a new better miner as I said and 99% of the cases it will work.

It could also be that Windows is using more than 900 MB for the graphics so you can't mine Eth anymore with this card.
jr. member
Activity: 49
Merit: 1
Hi everyone:


I should admit that I haven't been up to speed on crypto-mining lately. I built my rig a few years back and had it running and do its thing in my basement.

Here's my setup:

- I have 4 (genuine) GeForce GTX 1060 cards, each with 6GB of video RAM. (Bought them on newegg, so I'm assuming it's not a Chinese knockoff.)

- Mobo: ASUS PRIME Z270-A LGA 1151 Intel Z270 HDMI SATA 6Gb/s USB 3.1 ATX Motherboards - Intel

- CPU is: Intel(R) Core(TM) i3-6100 CPU @ 3.70GHz

- RAM: 8 GB - Single DDR4 2666 MT/s (PC4-21300) SR x8 DIMM 288-Pin Memory

- Drive: SSD -- 128 GB -- ADATA SU800 M.2 2280 128GB Ultimate 3D NAND

- Power supply: EVGA SuperNOVA 1000 G3, 80 Plus Gold 1000W


I was using the open-source ethminer to mine ETH for a couple years or so. Through the following command line:

Code:
ethminer.exe -P stratum://0xmy_eth_id.my_mining_rig:[email protected]:14444


So a few days back I got a notification from the mining pool that it hasn't received any good hashes. I checked my mining rig, rebooted it, and it was showing the following errors upon the start of the ethminer:

Code:
ethminer 0.19.0-alpha.0
Build: windows/release/msvc
No OpenCL platforms found
 i 06:27:07 main     Configured pool us2.ethermine.org:14444
 i 06:27:07 Selected pool us2.ethermine.org:14444
 i 06:27:07 Negotiation of EthereumStratum/2.0.0 failed. Trying another ...
 i 06:27:07 Stratum mode : EthereumStratum/1.0.0 (NiceHash)
 i 06:27:07 Established connection to us2.ethermine.org [172.65.230.90:14444]
 i 06:27:07 Spinning up miners...
cu 06:27:07 cuda-0   Using Pci Id : 01:00.0 GeForce GTX 1060 6GB (Compute 6.1) Memory : 6.00 GB
cu 06:27:07 cuda-1   Using Pci Id : 03:00.0 GeForce GTX 1060 6GB (Compute 6.1) Memory : 6.00 GB
cu 06:27:07 cuda-2   Using Pci Id : 04:00.0 GeForce GTX 1060 6GB (Compute 6.1) Memory : 6.00 GB
cu 06:27:07 cuda-3   Using Pci Id : 06:00.0 GeForce GTX 1060 6GB (Compute 6.1) Memory : 6.00 GB
 i 06:27:07 Authorized worker ___________________________
 i 06:27:08 Epoch : 516 Difficulty : 4.29 Gh
 i 06:27:08 Job: c956e546... us2.ethermine.org [172.65.230.90:14444]
 i 06:27:11 Job: 7c432b56... us2.ethermine.org [172.65.230.90:14444]
cu 06:27:11 cuda-3   Generating DAG + Light : 5.11 GB
cu 06:27:11 cuda-0   Generating DAG + Light : 5.11 GB
cu 06:27:11 cuda-3   Unexpected error CUDA error in func dev::eth::CUDAMiner::initEpoch_internal at line 123 out of memory on CUDA device 06:00.0
cu 06:27:11 cuda-3   Mining suspended ...
cu 06:27:11 cuda-2   Generating DAG + Light : 5.11 GB
cu 06:27:11 cuda-0   Unexpected error CUDA error in func dev::eth::CUDAMiner::initEpoch_internal at line 123 out of memory on CUDA device 01:00.0
cu 06:27:11 cuda-2   Unexpected error CUDA error in func dev::eth::CUDAMiner::initEpoch_internal at line 123 out of memory on CUDA device 04:00.0
cu 06:27:11 cuda-0   Mining suspended ...
cu 06:27:11 cuda-2   Mining suspended ...
cu 06:27:11 cuda-1   Generating DAG + Light : 5.11 GB
cu 06:27:11 cuda-1   Unexpected error CUDA error in func dev::eth::CUDAMiner::initEpoch_internal at line 123 out of memory on CUDA device 03:00.0
cu 06:27:11 cuda-1   Mining suspended ...

I did some search and found out that people were suggesting to increase the size of virtual memory. So I did that as well, raised it to 50GB:



rebooted, but the miner was still showing the same error.

Any idea what is going on there?




Jump to: