cuda-samples VS VkFFT

installing, building, and running WSL CUDA examples from https://github.com/nvidia/cuda-samples

After which nvcc should be accessible to new sessions, and you can build C++ cuda stuff like cuda-samples. Python packages like pytorch should also see CUDA and be able to use it.

I haven't found any note that these APIs do not support Windows, and it also seems that the memMapIPCDrv CUDA sample supports Windows.

I installed nvidia-cuda-toolkit and tried to build https://github.com/NVIDIA/cuda-samples but I'm getting stupid errors... it installed nvccat /usr/bin/nvcc and the samples expect /usr/local/cuda/bin/nvcc... symlinking it to that location and it dies with

Can build the cuda-samples, then you have a working nvcc.

I have installed cuda and cudnn, and was testing the installation with the cuda-samples, as the Arch Wiki suggested. But, I am not able to get samples like nbody, smokeparticles, Mandelbrot, etc. to run. Although devicequery works fine, and I get the expected output, so I think there is not a problem with my cuda installation.

Hi, I don't have much experience with Nvidia Jetsons. You can find some examples on GitHub (here https://github.com/NVIDIA/cuda-samples). You can find CUDA implementations of most functions on the internet though, you just have to look for the specific thing you are looking for. Cuda kernels are not platform specific, they should work on GPUs and embedded developer boards without problems as long as you respect the limits imposed by the "compute capability" of your device, you just have to compile your code using the right architecture flag. The biggest limit you have to deal with when developing for Jetson nano is the low amount of memory.

—> git clone https://github.com/NVIDIA/cuda-samples.git —> cd cuda-samples/Samples/1_Utilities/deviceQuery/ —> make —> ./deviceQuery (Result=pass?good) —> cd ~/ —> wget https://repo.anaconda.com/archive/Anaconda3-2021.11-Linux-x86_64.sh

is this the source code that u are talking about ? : https://github.com/NVIDIA/cuda-samples ? I dont see any CMakeLists.txt inside...

Hello, I am the creator of the VkFFT - GPU Fast Fourier Transform library for Vulkan/CUDA/HIP/OpenCL/Level Zero and Metal. In the latest update, I have added support for quad-precision double-double emulation for FFT calculation on most modern GPUs. I understand that modern ML is going in the opposite low-precision direction, but I still think that it may be useful to have this functionality at least for some prototyping and development of concepts.

Not quite what I asked for, but close enough for now...

https://github.com/DTolm/VkFFT/discussions/126

Source - I have made a somewhat functional programming-like FFT library (https://github.com/DTolm/VkFFT/tree/develop) which also operates on abstract data containers. Maybe it can be interesting to you from the algorithmic point of view.

VkFFT is a use-case I've heard of where Vulkan-Compute is faster than its Cuda and OpenCL counter-part: https://github.com/DTolm/VkFFT

Hello, I am the creator of the VkFFT - GPU Fast Fourier Transform library for Vulkan/CUDA/HIP/OpenCL and Level Zero. In the latest update, I have added support for Apple Metal API, which will allow VkFFT to run natively on modern Apple SoC. I have tested it on MacBook Pro with an M1 Pro 8c CPU/14c GPU SoC single precision on 1D batched FFT test of all systems from 2 to 4096. Achieved bandwidth is calculated as 2*system size divided by the time taken per FFT - minimum memory that has to be transferred between DRAM and GPU:

Another possible project to look at is https://github.com/DTolm/VkFFT