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

Failed to load CUDA fat binary #96

Closed
Jtag-Nathan opened this issue Nov 21, 2020 · 15 comments
Closed

Failed to load CUDA fat binary #96

Jtag-Nathan opened this issue Nov 21, 2020 · 15 comments

Comments

@Jtag-Nathan
Copy link

I am trying to mine some things with a DAG less than 4GB on some older GTX770 this is what I get when trying to run the miner.

20201121 15:47:53 T-Rex NVIDIA GPU miner v0.18.11 - [CUDA v11.10]
20201121 15:47:53 r.2c8851280f45
20201121 15:47:53
20201121 15:47:53 NVIDIA Driver v457.30
20201121 15:47:53 CUDA devices available: 1
20201121 15:47:53
20201121 15:47:53 WARN: DevFee 1% (ethash)
20201121 15:47:53
20201121 15:47:53 WARN: Fork has been requested at epoch 390. Algo will be changed to etchash.
20201121 15:47:53
20201121 15:47:53 URL : stratum+tcp://etc.2miners.com:1010
20201121 15:47:53 USER: 0x1f75eccd8fbddf057495b96669ac15f8e296c2cd
20201121 15:47:53 PASS: x
20201121 15:47:53 WRK : rig0
20201121 15:47:53
20201121 15:47:53 Starting on: etc.2miners.com:1010
20201121 15:47:53 ApiServer: HTTP server started on 0.0.0.0:4067
20201121 15:47:53 ---------------------------------------------------------
20201121 15:47:53 For control navigate to: http://192.168.137.147:4067/trex
20201121 15:47:53 ---------------------------------------------------------
20201121 15:47:53 ApiServer: Telnet server started on 0.0.0.0:4068
20201121 15:47:53 Using protocol: stratum1.
20201121 15:47:53 Authorizing...
20201121 15:47:53 ERROR: Can't initialize device [ID=0, GPU #0], failed to load CUDA fat binary
20201121 15:47:53 WARN: Miner is going to shutdown...
20201121 15:47:53 Authorized successfully.
20201121 15:47:53 Main loop finished. Cleaning up resources...
20201121 15:47:53 ApiServer: stopped listening on 0.0.0.0:4067
20201121 15:47:53 ApiServer: stopped listening on 0.0.0.0:4068
20201121 15:47:53 ethash epoch: 388, block: 11652390, diff: 1.192 = 8.59 Gh
20201121 15:47:54 WARN: NVML: can't get power for GPU #0, not supported
20201121 15:47:55 T-Rex finished.
20201121 15:47:55 WARN: WATCHDOG: T-Rex does not exist anymore, restarting...
20201121 15:47:58 T-Rex NVIDIA GPU miner v0.18.11 - [CUDA v11.10]
20201121 15:47:58 r.2c8851280f45
@trexminer
Copy link
Owner

GTX 770 is Compute Capability 3.0 according to https://developer.nvidia.com/cuda-gpus, but T-Rex only works with 5.0+. Sorry about that, your GPU is just too old. We definitely need to work on a better error message for similar cases.

@Jtag-Nathan
Copy link
Author

@trexminer Thanks ! not a problem at least I know its not toasted hardware !

@jonathanhatch
Copy link

I am having a similar problem. I am using A100 PCIe 4.0 cards:

20201202 16:51:10 Using protocol: stratum1. 20201202 16:51:10 Authorizing... 20201202 16:51:10 Authorized successfully. 20201202 16:51:10 ERROR: Can't initialize device [ID=0, GPU #0], failed to load CUDA fat binary 20201202 16:51:10 WARN: Miner is going to shutdown... 20201202 16:51:10 ethash epoch: 379, block: 11374288, diff: 4.00 Gh 20201202 16:51:11 ERROR: Can't initialize device [ID=1, GPU #1], failed to load CUDA fat binary 20201202 16:51:11 ERROR: Can't initialize device [ID=2, GPU #2], failed to load CUDA fat binary 20201202 16:51:11 ERROR: Can't initialize device [ID=3, GPU #3], failed to load CUDA fat binary 20201202 16:51:11 Main loop finished. Cleaning up resources... 20201202 16:51:11 ApiServer: stopped listening on 0.0.0.0:4067 20201202 16:51:11 ApiServer: stopped listening on 0.0.0.0:4068 20201202 16:51:11 ERROR: Can't initialize device [ID=4, GPU #4], failed to load CUDA fat binary 20201202 16:51:11 ERROR: Can't initialize device [ID=5, GPU #5], failed to load CUDA fat binary 20201202 16:51:11 ERROR: Can't initialize device [ID=6, GPU #6], failed to load CUDA fat binary 20201202 16:51:11 WARN: NVML: can't get fan speed for GPU #0, not supported 20201202 16:51:11 WARN: NVML: can't get fan speed for GPU #1, not supported 20201202 16:51:11 ERROR: Can't initialize device [ID=7, GPU #7], failed to load CUDA fat binary 20201202 16:51:12 T-Rex finished. 20201202 16:51:13 WARN: WATCHDOG: T-Rex does not exist anymore, restarting... 20201202 16:51:16 T-Rex NVIDIA GPU miner v0.19.1 - [CUDA v11.10] 20201202 16:51:16 r.99a7206c3590 20201202 16:51:16 20201202 16:51:16 NVIDIA Driver v455.45.01 20201202 16:51:16 CUDA devices available: 8 20201202 16:51:16 20201202 16:51:16 WARN: DevFee 1% (ethash) 20201202 16:51:16 20201202 16:51:16 URL : stratum+tcp://us2.ethermine.org:4444 20201202 16:51:16 USER: 20201202 16:51:16 PASS: x 20201202 16:51:16 WRK : rig0 20201202 16:51:16 20201202 16:51:16 Starting on: us2.ethermine.org:4444

@trexminer
Copy link
Owner

@jonathanhatch A100 is Compute Capability 8.0, we'll add support for it in the next version

@jonathanhatch
Copy link

@jonathanhatch A100 is Compute Capability 8.0, we'll add support for it in the next version

@trexminer Awesome. Do we have an ETA for that next version? I'm excited to get some ETH mining in while I still have the chance.

@trexminer
Copy link
Owner

trexminer commented Dec 4, 2020 •
edited
Loading

@jonathanhatch what OS are you running? I'll do a build for you before the next release as there's no ETA on it yet.

@jonathanhatch
Copy link

@jonathanhatch what OS are you running? I'll do a build for you before the next release as there's no ETA on it yet.

Honestly? That would be rad. I'm running a mixture of Ubuntu 18.04 LTS and 20.04 LTS machines.

@trexminer
Copy link
Owner

There you go: https://www.dropbox.com/s/lrbexnsr26q51un/t-rex-0.19.1-CC80-linux-cuda11.1.tar.gz?dl=1
Let me know if it works. And if it does, please save it locally as it won't stay in my Dropbox forever :)

@jonathanhatch
Copy link

I'll let you know as soon as I can get back into my ePDU to figure out how much power I have left on tap. I got myself locked out somehow. lolz

@alexanderthomas13
Copy link

having the same issue myself now after a GPU update with a card I thought was supported 1660ti

@trexminer
Copy link
Owner

@alexanderthomas13 1660ti is supported, plenty of people are running t-rex with it. Please provide a log similar to the one in the first message in this thread

@alexanderthomas13
Copy link

C:\Users\Alext\Downloads>t-rex.exe -a ethash -o stratum+tcp://us1.ethermine.org:4444 -u 0x41161eE534A0a499Cc0fa158701c343347099D10 -p x -w rig0
20210107 23:53:34 T-Rex NVIDIA GPU miner v0.19.5 - [CUDA v11.10 | Windows]
20210107 23:53:34 r.5f0b2f67355c
20210107 23:53:34
20210107 23:53:34 NVIDIA Driver v461.09
20210107 23:53:34 CUDA devices available: 2
20210107 23:53:34
20210107 23:53:34 WARN: DevFee 1% (ethash)
20210107 23:53:34
20210107 23:53:34 URL : stratum+tcp://us1.ethermine.org:4444
20210107 23:53:34 USER: 0x41161eE534A0a499Cc0fa158701c343347099D10
20210107 23:53:34 PASS: x
20210107 23:53:34 WRK : rig0
20210107 23:53:34
20210107 23:53:34 Starting on: us1.ethermine.org:4444
20210107 23:53:34 Using protocol: stratum1.
20210107 23:53:34 Authorizing...
20210107 23:53:34 ApiServer: HTTP server started on 0.0.0.0:4067
20210107 23:53:34 Authorized successfully.
20210107 23:53:34 -----------------------------------------------------
20210107 23:53:34 For control navigate to: http://192.168.0.6:4067/trex
20210107 23:53:34 -----------------------------------------------------
20210107 23:53:34 ApiServer: Telnet server started on 0.0.0.0:4068
20210107 23:53:34 ethash epoch: 387, block: 11612788, diff: 4.00 Gh
20210107 23:53:34 WARN: GPU #0(000200): Gigabyte GeForce GTX 1660 Ti, intensity set to 22
20210107 23:53:34 ERROR: Can't initialize device [ID=1, GPU #1], failed to load CUDA fat binary
20210107 23:53:34 WARN: Miner is going to shutdown...
20210107 23:53:34 Main loop finished. Cleaning up resources...
20210107 23:53:34 ApiServer: stopped listening on 0.0.0.0:4067
20210107 23:53:34 ApiServer: stopped listening on 0.0.0.0:4068
20210107 23:53:37 GPU #0: generating DAG 4.02 GB for epoch 387 ...
20210107 23:53:37 GPU #0: DAG generated [time: 471 ms], memory left: 1.02 GB
20210107 23:53:37 T-Rex finished.
20210107 23:53:38 WARN: WATCHDOG: T-Rex does not exist anymore, restarting...
20210107 23:53:40 T-Rex NVIDIA GPU miner v0.19.5 - [CUDA v11.10 | Windows]
20210107 23:53:40 r.5f0b2f67355c
20210107 23:53:40
20210107 23:53:40 NVIDIA Driver v461.09
20210107 23:53:40 CUDA devices available: 2
20210107 23:53:40
20210107 23:53:40 WARN: DevFee 1% (ethash)
20210107 23:53:40
20210107 23:53:40 URL : stratum+tcp://us1.ethermine.org:4444
20210107 23:53:40 USER: 0x41161eE534A0a499Cc0fa158701c343347099D10
20210107 23:53:40 PASS: x
20210107 23:53:40 WRK : rig0
20210107 23:53:40
20210107 23:53:40 Starting on: us1.ethermine.org:4444
20210107 23:53:40 Using protocol: stratum1.
20210107 23:53:40 Authorizing...
20210107 23:53:40 Authorized successfully.
20210107 23:53:40 ApiServer: HTTP server started on 0.0.0.0:4067
20210107 23:53:40 -----------------------------------------------------
20210107 23:53:40 For control navigate to: http://192.168.0.6:4067/trex
20210107 23:53:40 -----------------------------------------------------
20210107 23:53:40 ApiServer: Telnet server started on 0.0.0.0:4068
20210107 23:53:40 ethash epoch: 387, block: 11612788, diff: 4.00 Gh
20210107 23:53:40 WARN: GPU #0(000200): Gigabyte GeForce GTX 1660 Ti, intensity set to 22
20210107 23:53:40 ERROR: Can't initialize device [ID=1, GPU #1], failed to load CUDA fat binary
20210107 23:53:40 WARN: Miner is going to shutdown...
20210107 23:53:41 Main loop finished. Cleaning up resources...
20210107 23:53:41 ApiServer: stopped listening on 0.0.0.0:4067
20210107 23:53:41 ApiServer: stopped listening on 0.0.0.0:4068
20210107 23:53:43 GPU #0: generating DAG 4.02 GB for epoch 387 ...
20210107 23:53:43 GPU #0: DAG generated [time: 475 ms], memory left: 1.02 GB
20210107 23:53:44 T-Rex finished.
20210107 23:53:44 WARN: WATCHDOG: T-Rex does not exist anymore, restarting...

@trexminer
Copy link
Owner

Thanks. What is your second GPU that isn't showing up in the log?

@CDeLeon94
Copy link

@jonathanhatch Did you ever get mining on your a100? I'm curious what kind of hash rate you were able to get.

@jonathanhatch
Copy link

jonathanhatch commented Feb 12, 2021 via email

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
5 participants
@Jtag-Nathan @CDeLeon94 @jonathanhatch @trexminer @alexanderthomas13