Stack Overflow
Stack Overflow

Questions tagged [pycuda]

PyCUDA is the Python module which provides a comprehensive pythonic interface to the NVIDIA CUDA GPU computing environment.

PyCUDA provides a python module to access the NVIDIA CUDA driver API from within Python code.

The module includes interoperability with numpy, and comprehensive metaprogramming facilities for dynamically generating and JIT compiling CUDA code using Python.

417 questions
0
votes
1 answer

PyCUDA misaligned address cleanup failure

I am very new to CUDA programming and am starting off with PyCUDA to get the basics. I studied the tutorials and have run a couple of simple test codes. The tests used only 1D arrays. When I tried to run the following code with 2D arrays, I am…
Roshan Sam
  • 85
  • 9
0
votes
1 answer

Errors installing CUDA back-end (or cudarray) in Python in Windows 10

I am trying to follow the instructions to install CUDA (https://github.com/andersbll/cudarray). I am able to install the cudarray python package WITHOUT CUDA back-end following the instruction. The installation seems to be successful. When "pip…
KubiK888
  • 4,377
  • 14
  • 61
  • 115
0
votes
1 answer

PyCuda: Dereferencing Array Element Via Pointer in Cuda Kernel

I am using PyCuda to pass pairs of arrays to a cuda kernel via a pointer. The arrays are the output of a different kernel, so the data is already on the GPU. Within the kernel, I'm trying to access elements in each of the arrays to do a vector…
Alex Hall
  • 171
  • 11
0
votes
1 answer

Pycuda Concurrency

Why would the code from PyCuda KernelConcurrency Example not run faster in 'concurrent' mode? It seems like there should be enough resources on my GPU... what am I missing? Here is the output from the 'concurrent' version, with line 63…
Alex Hall
  • 171
  • 11
0
votes
1 answer

Pycuda Array Indexing with Threads & Blocks

I'm trying to write a cuda histogram function for use with Pycuda. The code seems to be iterating through more elements than are in the size of the array I'm passing in. To rule out errors in the bin computation, I created a very simple kernel where…
Alex Hall
  • 171
  • 11
0
votes
1 answer

Copying arrays with pyCuda from the GPU using pointers

I'm copying via pycuda some arrays on the GPU and then store the pointers to these arrays. How do I recuperate the data back? dist = np.zeros(numPoints).astype(np.float32) distAddress = [gpuarray.to_gpu(dist).ptr for i in range(100)] If I call the…
ZappaZ
  • 95
  • 1
  • 6
0
votes
1 answer

Unable to install pycuda on Mac OS X EL Capitan

I'm trying to install pycuda-2015.1.3 on my macbook pro. I've successfully installed CUDA, and I'm working with python 3.4. After entering terminal and going into the pycuda folder, I hit "sudo make" and get the following output: ctags -R src ||…
shiretzet
  • 21
  • 5
0
votes
0 answers

Installing pycuda on ubuntu 14.04

I want to install this package which uses pycuda on Ubuntu 14.04. (at this point version of pycuda don't matter) After following instructions from this page. I try to test the installation of pycuda with: ~/Downloads/pycuda-2015.1.3/test$ sudo…
Ricevind
  • 341
  • 1
  • 4
  • 15
0
votes
1 answer

CUDA runtime gpu initialization with theano

I am trying to parallelize my NN across two GPUs following https://github.com/uoguelph-mlrg/theano_multi_gpu. I have all the dependencies, but the cuda runtime initialization fails with the following message. ERROR (theano.sandbox.cuda): ERROR: Not…
baskaran
  • 25
  • 1
  • 7
0
votes
1 answer

How to use pyCUDA to broadcast via MPI?

Is there anyone, who ever sent CUDA arrays over MPI via most recent mpy4py ( and pyCUDA 2015.1.3)? To send an array, one must convert respective data type to the contiguous buffer. This conversion is done using the following lambda: to_buffer =…
Vast Academician
  • 357
  • 4
  • 12
0
votes
1 answer

Apparent CUDA magic

I'm using CUDA (in reality I'm using pyCUDA if the difference matters) and performing some computation over arrays. I'm launching a kernel with a grid of 320*600 threads. Inside the kernel I'm declaring two linear arrays of 20000 components…
Dargor
  • 623
  • 1
  • 4
  • 12
0
votes
1 answer

PyCUDA large nonuniform matrix operations

I am working with large, nonuniform matrices and am having problems with what I believe to be mismatching on the elements. In example.py, get_simulated_ipp() builds echo and tx, two linear arrays of size 250000 and 25000 respectively. The code also…
Nate Hilliard
  • 29
  • 4
0
votes
1 answer

TypeError: 'numpy.uint64' does not have the buffer interface

i am using pycuda and i write this program etat=np.zeros(XBLOCK * YBLOCK * XGRID * YGRID,dtype=np.uint) compteur_init=np.uint(0) clef_utilisateur=np.uint(SEED) config=clef_utilisateur compteur_init_gpu =…
SOCKet
  • 191
  • 1
  • 2
  • 15
0
votes
1 answer

getrs function of cuSolver over pycuda doesn't work properly

I'm trying to make a pycuda wrapper inspired by scikits-cuda library for some operations provided in the new cuSolver library of Nvidia. I want to solve a linear system of the form AX=B by LU factorization, to perform that first use the…
Miguel Diaz
  • 17
  • 5
0
votes
1 answer

Segmentation Fault in Pycuda using NVIDIA's cuSolver Library

i'm tryin to make a pycuda wrapper inspired by scikits-cuda library, for some operations provided in the new cuSolver library of Nvidia, first I need to perfom an LU factorization through cusolverDnSgetrf() op. but before that I need the…
Miguel Diaz
  • 17
  • 5
1 2 3
…
27 28