hashcat fails to start on Windows 11 - error on OpenCLOn12.dll

[Itismo]

did you try to uninstall all drivers and install the vendor driver as suggested in the github issue ?

does nvidia officially support windows 11 already... Probably it already does... but is somebody else able to test on windows 11 if hashcat works for them ?

(note: this problem was already reported here: https://github.com/hashcat/hashcat/issues/3216)

---

update:
yeah, it seems that there is support for windows 11 already for your GPU:  https://www.nvidia.com/Download/driverRe...7099/en-us , again I would suggest to uninstall this microsoft drivers, maybe use driver fusion or DDU (Display Driver Uninstaller) or similar tools to get rid of this microsoft OpenCLOn12.dll runtime and/or opencl implementation.
You probably only need the nvidia driver that ships with all runtimes needed for hashcat

Phil, your help is much appreciated and thank you so much for the simple but overlooked hint to uninstall the Microsoft OpenCL and OpenGL Compatibilty Pack that shipped with Windows, after that I am able to get hashcat to start! I am wondering if there is any downside to uninstalling that?

Now, hashcat finds 3 GPUs not sure why there is a duplicate for the Nvidia, and when I leave it without using -d, I see the GPU utilization increasing on the Intel Iris card AND also the memory usage is increasing on the Nvidia card (but not GPU processing, which confuses me as to why GPU processing is used on Intel card and memory used from both Nvidia and Intel.

If I specify the device with CUDA API using -d 1 (OpenCL API (OpenCL 3.0 CUDA 11.6.110) - Platform #1 NVIDIA GeForce RTX 3050 Laptop GPU, 3335/4095 MB, 16MCU), only memory usage increases on the Nvidia GPU but NOT the 3D utilization, the Intel Iris is sleeping, no issues.

If I specify -d 2 (OpenCL API (OpenCL 3.0 ) NVIDIA GeForce RTX 3050 Laptop GPU)  to choose the OpenCL API, it stays forvever on "Initializing backend runtime for device #2. Please be patient..."

If I specify -d 3 (Device #3: Intel(R) Iris(R) Xe Graphics, 3168/6451 MB (1612 MB allocatable), 96MCU).



CUDA API (CUDA 11.6)
====================
* Device #1: NVIDIA GeForce RTX 3050 Laptop GPU, 3335/4095 MB, 16MCU
OpenCL API (OpenCL 3.0 CUDA 11.6.110) - Platform #1 [NVIDIA Corporation]
========================================================================
* Device #2: NVIDIA GeForce RTX 3050 Laptop GPU, skipped
OpenCL API (OpenCL 3.0 ) - Platform #2 [Intel(R) Corporation]
=============================================================
* Device #3: Intel(R) Iris(R) Xe Graphics, 3168/6451 MB (1612 MB allocatable), 96MCU

Intel Iris GPU  96% usage


Nvidia GPU 0% usage but 3 GB memory usage