CUDA Issues with Nvidia 3080s (hc v6.2.5)

[rumham]

Hey all, I recently built a new rig and am running into an issue with what appears to be CUDA from my hashcat output. Pretty strange as I believe I've installed the proper nvidia drivers and latest cuda sdk toolkit. I'm sure I'm missing something simple here... Below is some output from my system. 

GPU info:
4x Gigabyte NVIDIA GeForce RTX 3080

Hashcat Version = v6.2.5

uname -a

Code:

Linux loki 5.15.0-43-generic #46-Ubuntu SMP Tue Jul 12 10:30:17 UTC 2022 x86_64 x86_64 x86_64 GNU/Linux

Hashcat output testing benchmark against NTLMv2. 

Code:

hashcat -m 5600 -b

hashcat (v6.2.5) starting in benchmark mode

Benchmarking uses hand-optimized kernel code by default.
You can use it in your cracking session by setting the -O option.
Note: Using optimized kernel code limits the maximum supported password length.
To disable the optimized kernel code in benchmark mode, use the -w option.

Successfully initialized NVIDIA CUDA library.

Failed to initialize NVIDIA RTC library.

* Device #1: CUDA SDK Toolkit not installed or incorrectly installed.
            CUDA SDK Toolkit required for proper device support and utilization.
            Falling back to OpenCL runtime.

Successfully initialized NVIDIA CUDA library.

Failed to initialize NVIDIA RTC library.

* Device #2: CUDA SDK Toolkit not installed or incorrectly installed.
            CUDA SDK Toolkit required for proper device support and utilization.
            Falling back to OpenCL runtime.

Successfully initialized NVIDIA CUDA library.

Failed to initialize NVIDIA RTC library.

* Device #3: CUDA SDK Toolkit not installed or incorrectly installed.
            CUDA SDK Toolkit required for proper device support and utilization.
            Falling back to OpenCL runtime.

Successfully initialized NVIDIA CUDA library.

Failed to initialize NVIDIA RTC library.

* Device #4: CUDA SDK Toolkit not installed or incorrectly installed.
            CUDA SDK Toolkit required for proper device support and utilization.
            Falling back to OpenCL runtime.

OpenCL API (OpenCL 3.0 CUDA 11.7.99) - Platform #1 [NVIDIA Corporation]
=======================================================================
* Device #1: NVIDIA GeForce RTX 3080, 9792/10016 MB (2504 MB allocatable), 68MCU
* Device #2: NVIDIA GeForce RTX 3080, 9792/10018 MB (2504 MB allocatable), 68MCU
* Device #3: NVIDIA GeForce RTX 3080, 9792/10018 MB (2504 MB allocatable), 68MCU
* Device #4: NVIDIA GeForce RTX 3080, 9792/10018 MB (2504 MB allocatable), 68MCU

Benchmark relevant options:
===========================
* --optimized-kernel-enable

----------------------------
* Hash-Mode 5600 (NetNTLMv2)
----------------------------

Speed.#1.........:  4148.3 MH/s (68.66ms) @ Accel:64 Loops:256 Thr:256 Vec:1
Speed.#2.........:  4141.1 MH/s (68.77ms) @ Accel:64 Loops:256 Thr:256 Vec:1
Speed.#3.........:  4094.2 MH/s (69.55ms) @ Accel:64 Loops:256 Thr:256 Vec:1
Speed.#4.........:  4138.1 MH/s (68.80ms) @ Accel:64 Loops:256 Thr:256 Vec:1
Speed.#*.........: 16521.7 MH/s

Started: Fri Jul 29 19:50:13 2022

Stopped: Fri Jul 29 19:50:27 2022

nvidia-smi output

Code:

Fri Jul 29 19:49:23 2022     
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 515.57      Driver Version: 515.57      CUDA Version: 11.7    |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|        Memory-Usage | GPU-Util  Compute M. |
|                              |                      |              MIG M. |
|===============================+======================+======================|
|  0  NVIDIA GeForce ...  Off  | 00000000:05:00.0 Off |                  N/A |
|  0%  52C    P0    99W / 370W |      0MiB / 10240MiB |      0%      Default |
|                              |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
|  1  NVIDIA GeForce ...  Off  | 00000000:06:00.0 Off |                  N/A |
| 55%  59C    P0  108W / 370W |      0MiB / 10240MiB |      0%      Default |
|                              |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
|  2  NVIDIA GeForce ...  Off  | 00000000:09:00.0 Off |                  N/A |
| 63%  58C    P0  111W / 370W |      0MiB / 10240MiB |      0%      Default |
|                              |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
|  3  NVIDIA GeForce ...  Off  | 00000000:0A:00.0 Off |                  N/A |
|  0%  57C    P0    91W / 370W |      0MiB / 10240MiB |      0%      Default |
|                              |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                             
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU  GI  CI        PID  Type  Process name                  GPU Memory |
|        ID  ID                                                  Usage      |
|=============================================================================|
|  No running processes found                                                |
+-----------------------------------------------------------------------------+

Happy to provide more info/context as requested.

[Chick3nman]

Did you follow the post installation steps of adding everything to the PATH/ENV?

https://docs.nvidia.com/cuda/cuda-instal...on-actions

[rumham]

Hey Chick3nman, thanks for the assist there. That fixed the issue above. Now I'm running into an additional problem. Seems hashcat is hanging directly after listing the devices and right before starting a benchmark:

Code:

hashcat (v5.1.0) starting in benchmark mode...

Benchmarking uses hand-optimized kernel code by default.
You can use it in your cracking session by setting the -O option.
Note: Using optimized kernel code limits the maximum supported password length.
To disable the optimized kernel code in benchmark mode, use the -w option.

OpenCL Platform #1: NVIDIA Corporation
======================================
* Device #1: NVIDIA GeForce RTX 3080, 2504/10016 MB allocatable, 68MCU
* Device #2: NVIDIA GeForce RTX 3080, 2504/10018 MB allocatable, 68MCU
* Device #3: NVIDIA GeForce RTX 3080, 2504/10018 MB allocatable, 68MCU
* Device #4: NVIDIA GeForce RTX 3080, 2504/10018 MB allocatable, 68MCU

Benchmark relevant options:
===========================
* --optimized-kernel-enable

Hashmode: 0 - MD5

[Snoopy]

hashcat (v5.1.0) starting in benchmark mode...

seems you are invoking an old installed version of hashcat (distribution?) instead of your downloaded/self compiled one

try running directly in your 6.2.5 directory with prepended ./

[rumham]

Okay so I went back and ran with v6.2.5 and am getting the same CUDA issue again. Not sure what the difference being with v.5.1.0 and v.6.2.5.

I did make sure to double check that I followed this exactly https://docs.nvidia.com/cuda/cuda-instal...on-actions

[Snoopy]

Okay so I went back and ran with v6.2.5 and am getting the same CUDA issue again. Not sure what the difference being with v.5.1.0 and v.6.2.5.

I did make sure to double check that I followed this exactly https://docs.nvidia.com/cuda/cuda-instal...on-actions

can you please show your output of hashcat -I
benchmark tries all possible devices and it seem it starts with opencl on your cards instead of cuda

Code:

OpenCL Platform #1: NVIDIA Corporation

depending on the opencl version (the old 1.2 takes really long for initializing)