Questions Tagged [Cuda]
Ask Question Cuda (Compute Unified Device Architecture) Is a Parallel Computing Platform and Programming Model for Nvidia Gpus (Graphics Processing Units)...
![Questions Tagged [Cuda]](https://www.alledaags.nl/uploads/questions-tagged-cuda_1200.jpg)
CUDA (Compute Unified Device Architecture) is a parallel computing platform and programming model for NVIDIA GPUs (Graphics Processing Units). CUDA provides an interface to NVIDIA GPUs through a variety of programming languages, libraries, and APIs.
13,673
questions
3
votes
1
answer
17
views
Cuda AtomicAdd for int3
In Cuda AtomicAdd for double can be implemented using a while loop and AtomicCAS operation.
But how could I implement an atomic add for type int3 efficiently?
-3
votes
0
answers
12
views
Problem with argsort function when using guvectorize with cuda target
please i tested the following code using googlecolab and i get the following error:
Failed in nopython mode pipeline (step: nopython frontend)
Use of unsupported NumPy function ‘numpy.argsort’ or ...
-3
votes
0
answers
36
views
Tensorflow does not see GPU on Jupyter notebook [closed]
I have installed tensorflow and tensorflow-gpu on my project's anaconda environment as per the instalation guide:
When running this test line on terminal, my GPU ...
1
vote
0
answers
31
views
Concurrent Writing CUDA
I am new to CUDA and I am facing a problem with a basic projection kernel. What I am trying to do is to project a 3D point cloud into a 2D image. In case multiple points project to the same pixel, ...
0
votes
0
answers
27
views
Running an Nsight Systems report python script independently
I've tweaked a copy of one of the Nsight Systems report scripts (gpukernsum), and I now want to run it myself. So, I write:
./gpukernsum.py report.sqlite
This doesn't work; I get:
ERROR: Script '...
-2
votes
0
answers
35
views
Error applying cuda target to numba @guvectorize decorator [closed]
High
I have written this code to solve the test case elliptic equation using the Successive Over Relaxation (SOR) method.
I faced
import numpy,xarray
import numpy as np
from numba import vectorize, ...
5
votes
2
answers
61
views
Should CUDA stream be waited to be complete even if the output data are to be sent to OpenGL instead of CPU?
This is a general question, and although I use OpenCV as a framework, the question is broader than OpenCV's realm.
I am developing an image processing tool that will effectively get image from a ...
0
votes
0
answers
58
views
Implement convolution as a multiplication between matrices in Cuda
I am trying, for a university project, to implement the convolution on the Tensor Cores of a RTX2060.
We are therefore trying to map the convolution on a multiplication between matrices, looking ...
1
vote
1
answer
35
views
Getting total execution time of all kernels on a CUDA stream
I know how to time the execution of one CUDA kernel using CUDA events, which is great for simple cases. But in the real world, an algorithm is often made up of a series of kernels (CUB::...
0
votes
0
answers
58
views
ctypes could not access the 2D array
cudatest.py
import numpy as np
import ctypes
from ctypes import *
from numpy.ctypeslib import ndpointer
def get_cuda_square():
dll = ctypes.windll.LoadLibrary("C:\\Users\\Ali\\Desktop\\...
2
votes
1
answer
54
views
nppi resize function with 3 channels getting strange output
I'm getting a strange error when using nppi geometry transform functions from nppi cuda libraries. The code is here:
#include <nppi.h>
#include <nppi_geometry_transforms.h>
#include <...
0
votes
1
answer
43
views
Using dynamic parallelism results in 30x worse performance
Note: I don't have my computer and GPU with me so this me typing from memory. I timed this and compiled it correctly so ignore any odd typos should they exist.
I don't know if the overhead of what I'm ...
-2
votes
0
answers
17
views
Even after installing Cuda toolkit 11.7, nvcc -V release shows 10.1 [duplicate]
I had installed an older cuda-toolkit(10.1) via the command sudo apt install nvidia-cuda-toolkit
Since 10.1 is not supported for newer GPU models like NVIDIA GeForce RTX, I've got to install the 11.7 ...
-2
votes
0
answers
23
views
How to use nvjpeg instead of JPEG Lib？ [closed]
Jpeglib functions are heavily used in my project.I want to use GPU to speed up the process of JPEG decompression.I think I may need to encapsulate the related functions of nvjpeg instead of JPEG Lib....
1
vote
0
answers
34
views
Are load and store operations in shared memory atomic?
I'm trying to figure out whether load and store operations on primitive types are atomics when we load/store from shared memory in CUDA.
On the one hand, it seems that any load/store is compiled to ...
Sophia analyzes international trade, startup ecosystems, retail transformation, and supply chain logistics for modern digital publications.
