Stack Overflow
Stack Overflow

Questions tagged [gpu]

Acronym for "Graphics Processing Unit". For programming traditional graphical applications, see the tag entry for "graphics programming". For general-purpose programming using GPUs, see the tag entry for "gpgpu". For specific GPU programming technologies, see the popular tag entries for "opencl", "cuda" and "thrust".

Acronym for "Graphics Processing Unit". For programming traditional graphical applications, see the tag entry for . For general-purpose programming using GPUs, see the tag entry for . For specific GPU programming technologies, see the popular tag entries for , and .

More information on GPU at http://en.wikipedia.org/wiki/Graphics_processing_unit

8854 questions
27
votes
2 answers

How is CUDA memory managed?

When I run my CUDA program which allocates only a small amount of global memory (below 20 M), I got a "out of memory" error. (From other people's posts, I think the problem is related to memory fragmentation) I try to understand this problem, and…
xhe8
  • 419
  • 1
  • 5
  • 13
27
votes
5 answers

how to programmatically determine available GPU memory with tensorflow?

For a vector quantization (k-means) program I like to know the amount of available memory on the present GPU (if there is one). This is needed to choose an optimal batch size in order to have as few batches as possible to run over the complete data…
Barden
  • 1,020
  • 1
  • 10
  • 17
27
votes
9 answers

Cuda: library nvvm not found

I am trying to run the code below but an error is reported: NvvmSupportError: libNVVM cannot be found. Do conda install cudatoolkit: library nvvm not found My development environment is: Ubuntu 17.04, Spyder/Python3.5 and I have installed via conda…
Helton Maia
  • 271
  • 1
  • 4
  • 5
27
votes
2 answers

Very low GPU usage during training in Tensorflow

I am trying to train a simple multi-layer perceptron for a 10-class image classification task, which is a part of the assignment for the Udacity Deep-Learning course. To be more precise, the task is to classify letters rendered from various fonts…
Aleksei Petrenko
  • 6,698
  • 10
  • 53
  • 87
27
votes
9 answers

Using CUDA with Visual Studio 2017

I'm trying to install CUDA, but I get a message saying "No supported version of visual studio was found". I think that this is because I am using Visual Studio 2017 (Community), and CUDA currently only supports up to Visual Studio 2015.…
acernine
  • 729
  • 2
  • 6
  • 16
27
votes
2 answers

How to interpret TensorFlow output?

How do I interpret the TensorFlow output for building and executing computational graphs on GPGPUs? Given the following command that executes an arbitrary tensorflow script using the python API. python3 tensorflow_test.py > out The first part…
Alexander R Johansen
  • 2,737
  • 3
  • 18
  • 23
27
votes
2 answers

SLI for multiple GPUs

I am new to CUDA programming, and I am working on a problem that requires multiple GPUs in one machine. I understand that for better graphics programming multiple GPUs need to be combined via SLI. However, for CUDA programming do I need to combine…
gmemon
  • 2,573
  • 5
  • 32
  • 37
27
votes
8 answers

Why can't libcudart.so.4 be found when compiling the CUDA samples under Ubuntu?

I'm trying to get my Cuda SDK samples running, but I get the following error: ./bandwidthTest: error while loading shared libraries: libcudart.so.4: cannot open shared object file: No such file or directory Why can…
Alex
  • 503
  • 2
  • 6
  • 14
26
votes
1 answer

nvidia-smi: GPU's full name not displayed properly

On a Linux server, nvidia-smi can show GPU name, as well as other useful information. However, the full name is not displayed properly. How can I know the full name of GPUs?
fatpanda2049
  • 483
  • 1
  • 4
  • 9
26
votes
4 answers

GPU accelerated math in the browser

I'm starting a project for browsers which requires some complex data processing. The algorithm I'm using is 50-100x faster when accelerated with GPU. I could use JavaScript, Flash or other technologies with the browser. Is there any way I can access…
micho
  • 2,196
  • 2
  • 21
  • 26
26
votes
3 answers

Install Cuda without root

I know that I can install Cuda with the following: wget http://developer.download.nvidia.com/compute/cuda/7_0/Prod/local_installers/cuda_7.0.28_linux.run chmod +x cuda_7.0.28_linux.run ./cuda_7.0.28_linux.run -extract=`pwd`/nvidia_installers cd…
user200340
  • 3,301
  • 13
  • 52
  • 74
26
votes
5 answers

Tensorflow: Cuda compute capability 3.0. The minimum required Cuda capability is 3.5

I am installing tensorflow from source (documentation). Cuda driver version: nvcc: NVIDIA (R) Cuda compiler driver Cuda compilation tools, release 7.5, V7.5.17 When I ran the following command : bazel-bin/tensorflow/cc/tutorials_example_trainer…
Abhijay Ghildyal
  • 4,044
  • 6
  • 33
  • 54
26
votes
1 answer

Tensorflow multiple sessions with multiple GPUs

I have a workstation with 2 GPUs and I am trying to run multiple tensorflow jobs at the same time, so I can train more than one model at once, etc. For example, I've tried to separate the sessions into different resources via the python API using in…
j314erre
  • 2,737
  • 2
  • 19
  • 26
25
votes
7 answers

OpenCL - How to I query for a device's SIMD width?

In CUDA, there is a concept of a warp, which is defined as the maximum number of threads that can execute the same instruction simultaneously within a single processing element. For NVIDIA, this warp size is 32 for all of their cards currently on…
Jonathan DeCarlo
  • 2,798
  • 1
  • 20
  • 24
25
votes
3 answers

How to run tensorflow with gpu support in docker-compose?

I want to create some neural network in tensorflow 2.x that trains on a GPU and I want to set up all the necessary infrastructure inside a docker-compose network (assuming that this is actually possible for now). As far as I know, in order to train…
Kevin Südmersen
  • 883
  • 2
  • 14
  • 24
1 2 3
…
99 100