Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

GPU #0: FATAL: failed to allocate device memory for long state #31

Open
ghost opened this issue Nov 12, 2017 · 26 comments
Open

GPU #0: FATAL: failed to allocate device memory for long state #31

ghost opened this issue Nov 12, 2017 · 26 comments

Comments

@ghost
Copy link

ghost commented Nov 12, 2017 •
edited by ghost
Loading

i am running a evga GTX 750 Ti i keep getting this error, i think it has something to do with the .bat file but not sure, i think i have to set or change the blocks and threads but not sure, can anyone help me.
tsiv-xmr
[2017-11-11 19:33:35] Using JSON-RPC 2.0
[2017-11-11 19:33:35] 1 miner threads started, using 'cryptonight' algorithm.
[2017-11-11 19:33:35] Starting Stratum on stratum+tcp://xmr-us-east1.nanopool.org:14444
[2017-11-11 19:33:35] GPU #0: GeForce GTX 750 Ti (5 SMX), using 40 blocks of 8 threads
[2017-11-11 19:33:35] GPU #0: FATAL: failed to allocate device memory for long state
@mountassir
Copy link

Same issue here with GTX1070.

@auroracoin
Copy link

auroracoin commented Nov 12, 2017 •
edited
Loading

@darrellschwass -l 8x30 is what you should set your parameters for the 750ti. I get 280 h/s. But this error I don't know about. Sorry. I'm actually using a different app. but will be trying tsiv app soon. :-)

@ghost
Copy link
Author

ghost commented Nov 14, 2017 via email

@will2022
Copy link

will2022 commented Dec 1, 2017

I also have that issue, i'm mining on Nvidia GTX 1080 8 GB Strix. I think cpu mining is the alternate solution on my Ryzen 7

@dstana
Copy link

dstana commented Dec 6, 2017

Same issue GTX 1060.

@Honowski
Copy link

Honowski commented Dec 7, 2017

I've got the same error from both ccminer-cryptonight and tsiv-xmr. I've got GTX1060s

@ghost
Copy link

ghost commented Dec 8, 2017

I am also getting this issue when using the updated 388.13 Drivers for NVidia, but I was able to resolve it by downgrading drivers to 372.90, although I am not sure what is causing the probelm.

@Dutchy-79
Copy link

Same problem here
[2017-12-08 09:51:15] Using JSON-RPC 2.0
[2017-12-08 09:51:15] 1 miner threads started, using 'cryptonight' algorithm.
[2017-12-08 09:51:15] Starting Stratum on stratum+tcp://monerohash.com:3333
[2017-12-08 09:51:15] GPU #0: GeForce GTX 960 (8 SMX), using 40 blocks of 8 threads
[2017-12-08 09:51:15] GPU #0: FATAL: failed to allocate device memory for long state

@Azanigin
Copy link

1080ti same problem :(

@NoSubstitute
Copy link

1070 same problem, drivers v388.59

@Jertheunited
Copy link

Jertheunited commented Dec 12, 2017 •
edited
Loading

Same with GTX1070ti v388.59
image

@ghost
Copy link
Author

ghost commented Dec 13, 2017 via email

@Honowski
Copy link

I found the solution. Download https://github.com/fireice-uk/xmr-stak. Works on windows and Linux on both AMD and Nvidia cards. It also utilizes your CPU. I haven't had any issues with it.

@yonicsurny
Copy link

Same issue :-(
[2017-12-21 01:02:30] Using JSON-RPC 2.0
[2017-12-21 01:02:30] 1 miner threads started, using 'cryptonight' algorithm.
[2017-12-21 01:02:30] Starting Stratum on stratum+tcp://xmr.pool.minergate.com:45560
[2017-12-21 01:02:30] GPU #0: GeForce GTX 970 (13 SMX), using 40 blocks of 8 threads
[2017-12-21 01:02:30] GPU #0: Warning: block count 40 is not a multiple of SMX count 13.
[2017-12-21 01:02:31] Pool set diff to 1063
[2017-12-21 01:02:31] Stratum detected new block
[2017-12-21 01:02:34] GPU #0: FATAL: failed to allocate device memory for long state

@tousifhasanzai
Copy link

downgrade your cuda from 9 to 8

@Kastor83
Copy link

Same issue on 1070Ti and Titan V with current drivers.

Can you explain 'downgrade cuda from 9 to 8' please?

@SilverCory
Copy link

@Kastor83 https://developer.nvidia.com/cuda-80-ga2-download-archive
I'm just doing this I'll edit this comment if it works.

@MADeveloper
Copy link

Had this error on Win10, GTX 1060 with Cuda 9. I installed Cuda 8 and it corrected it. The miner is now working

@SilverCory
Copy link

SilverCory commented Dec 30, 2017 •
edited
Loading

I discovered this maintained fork. https://github.com/KlausT/ccminer-cryptonight

It has cuda 9 support. I had trouble installing cuda 8.

@ArtemkaKun
Copy link

Hi gues! If You have the Fatal Error, do next steps:

  1. Install https://www.microsoft.com/pl-pl/download/details.aspx?id=48145
  2. Install https://www.visualstudio.com/downloads/ (CUDA NEED IT!!! WITHOUT VS 2013(or better) CUDA DOESN'T INSTALL!!!)
  3. Install https://developer.nvidia.com/cuda-80-ga2-download-archive
  4. If You have troubles with step 3 - download latest drivers for your card and install it with CLEAR INSTALLATION
  5. Reboot your PC

  6. PROFIT!111!

@Kahless77
Copy link

Im downloaded and installed CUDA8, and this ccminer its just not working, same error message before CUDA8, any trick?

@ArtemkaKun
Copy link

@Kahless77 , try to install CUDA 9

@ash-predictiva
Copy link

Same Issue, tried all suggestions, nothing!

Installed VS
Installed CUDA 8
Installed CUDA 9
Updated Drivers

Nothing works!!

@ash-predictiva
Copy link

Here's the solution after long trials:

1- Download VS
2- Download CUDA 9
3- Download this version of CCminer (ccminer-x64-2.2.5) from here
https://github.com/tpruvot/ccminer/releases

4- Format your patch as follows

ccminer-x64.exe -a cryptonight -o YOUR POOL ADDRESS -u YOUR WALLET ADDRESS -p x

@auroracoin
Copy link

If your running windows, get the latest updates. May help.

@milnomada
Copy link

Hi,

I'm using:

  • cuda-6.5
  • nvidia geforce GS 9300M
  • compute_30 settings

The following command works:

./ccminer -a monero -o stratum+tcp://xmr-eu1.nanopool.org:14444 -u x -p {addr} -l 4x20 --bfactor=2

Having a look at the settings:

-l, --launch=CONFIG   launch config for the Cryptonight kernel.
                        a comma separated list of values in form of
                        AxB where A is the number of threads to run in
                        each thread block and B is the number of thread
                        blocks to launch. If less values than devices in use
                        are provided, the last value will be used for
                        the remaining devices. If you don't need to vary the
                        value between devices, you can just enter a single value
                        and it will be used for all devices. (default: 8x40)
--bfactor=X    Enables running the Cryptonight kernel in smaller pieces.
                        The kernel will be run in 2^X parts according to bfactor,

to tweak these values to smaller numbers to fit your card features.

[2021-05-25 17:32:16] GPU #0: GeForce 9300M GS, 93253.05 H/s

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests