Réponse apportée
Matlab GPU processing: Nvidia Quadro k2200 vs Geforce GTX 970
I cannot speak from experience. The Quadro series are intended for high-end visualisation, and GeForce for high speed graphics (...

plus de 10 ans il y a | 0

| A accepté

Réponse apportée
How do I use arrayfun on GPU when the size of the output array doesn't equal the size (of some) input arrays (Error using gpuArray/arrayfun
|gpuArray| |arrayfun| is different from the CPU version, which is effectively just a syntactic convenience to avoid writing |for...

plus de 10 ans il y a | 0

Réponse apportée
Manipulating arrays within GPU arrayfun
The general rule is that you cannot manipulate matrices or vectors inside a |gpuArray| |arrayfun| function, only scalar operatio...

plus de 10 ans il y a | 1

| A accepté

Réponse apportée
Matrix algebra very slow on GPU
It might be worth answering this question for posterity. The questioner it seems was testing at least the linear solves with ...

plus de 10 ans il y a | 0

Réponse apportée
Solve large linear equation on GPU
If you look online, or perhaps read a book like <https://www.amazon.co.uk/dp/1421407949/ref=pd_lpo_sbs_dp_ss_1/277-0329359-50611...

plus de 10 ans il y a | 1

| A accepté

Réponse apportée
Parallel Computing - use 8 cores and 2 video cards
A couple of questions answered: * Your graphics card is optimised for single precision compute. Make sure your data is in sin...

plus de 10 ans il y a | 0

| A accepté

Réponse apportée
GPU performance with short vectors
Computation in a GPU core is significantly slower than in a modern CPU core. It makes up for that by having a lot of them - thou...

plus de 10 ans il y a | 0

| A accepté

Réponse apportée
How to increase the speed of submatrix data transfer using gpuArray rather than CPU?
I think the problem is intrinsic. Even on my Kepler K20c the GPU is slower for this case. This is a memory intensive non-coheren...

plus de 10 ans il y a | 0

Réponse apportée
Efficient training of LSTM network with GPU
To get good performance out of the GPU, you need to give it a lot of data to process. Your best bet is to vectorize your code to...

plus de 10 ans il y a | 0

| A accepté

Réponse apportée
How do I use @mldivide together with pagefun on the GPU?
Your systems are rectangular. You can only solve square systems with |pagefun|, sorry.

plus de 10 ans il y a | 1

| A accepté

Réponse apportée
How to use multiple GPUs in distributed computing server?
Try |gpuDeviceCount| to determine how many GPUs a worker has access to, and then some manipulation of |labindex| to assign devic...

plus de 10 ans il y a | 1

Réponse apportée
How does one apply max(0,z) element-wise in a vectorized way (apply ReLu linear rectifier element-wise)
Not quite sure whether your point is clear because this is exactly how the two-argument form of |max| works - element-wise. ...

plus de 10 ans il y a | 0

Réponse apportée
Serious Bug in GPU Accelerated Interp1? -> EDIT: Present in R2015a, Fixed in R2016a
I can reproduce your slowdown, but I can't explain it since there have been no changes to the |interp| functions in recent relea...

plus de 10 ans il y a | 0

Réponse apportée
circshift slower on GPU
Yes, |circshift| is a bit slow on the GPU particularly when it has to shift both rowwise and columnwise, because that means it h...

plus de 10 ans il y a | 3

| A accepté

Réponse apportée
Parallel looped interp1 on GPU
The best way to parallelize multiple 1D interpolations is to use 2D interpolation, and just set the Y interp point to |(1:M)'|, ...

plus de 10 ans il y a | 1

| A accepté

plus de 10 ans il y a | 0

| A accepté

Réponse apportée
Mexcuda Error on Windows 10 (Matlab 2015 b)
As you can see from the first line of your output, your chosen host compiler is not supported by the CUDA compiler (nvcc) and yo...

plus de 10 ans il y a | 0

Réponse apportée
My code in Matlab takes longer time on Gpu compare with Cpu. my Gpu device is (GeForce 980 ti), could you some one run it on his/her gpu to see problem isrelated to code or my GPU hardwarte? (here is my code)
The problem is with your code. You need to vectorize the |mean| operation. At the moment you are calling thousands of kernels, s...

plus de 10 ans il y a | 0

Réponse apportée
I run the example "Stencil Operations on a GPU" in matlab command. I got following output. I don't have any advantage in using of Arrayfun on GPU compared with simple GPU array. my GPU device is GeForce GTX 980 Ti. is it reasonable?
Yes, it's perfectly reasonable.

plus de 10 ans il y a | 0

| A accepté

Réponse apportée
How can I set shared memory configuration for NVIDIA graphics card to increase L1 cache?
You can't, but you can write a mex function to do it, that will affect your next CUDAKernel call. MEX functions do not initia...

plus de 10 ans il y a | 0

Réponse apportée
is it normal that the computation time on GPU for the first time be longer than the running it for more than 2 times? (run the code for first time on GPU and also after reset the device takes more time but after 2 or 3 times it shows the short time))
Firstly, yes, it's perfectly normal. The first time you call a GPU function, the GPU libraries are loaded into MATLAB, which tak...

plus de 10 ans il y a | 2

| A accepté

plus de 10 ans il y a | 0

| A accepté

Réponse apportée
How to rewrite parfor loop using arrayfun to compute on GPU?
Seems like you're just trying to add together all the elements of A in every permutation, with the proviso that k > j > i. So re...

plus de 10 ans il y a | 0

Réponse apportée
In-place operation at gpuarray
Do it inside a function. On the command line MATLAB cannot operate on the arguments to functions in-place, because (to cite one ...

plus de 10 ans il y a | 2

Réponse apportée
Why for GPU is slower than CPU for this code? Is it because of sparsity or because of "for" loop
What you have here is a highly serial algorithm, accumulating small amounts of data inside a large array. To vectorize, perhaps ...

plus de 10 ans il y a | 0

Réponse apportée
Having problem In MatConvNet to Compiling the cuDNN support.
This appears to be a bug in MatConvNet's |vl_compilenn| function preventing it from working when your cudnn include path contain...

plus de 10 ans il y a | 1

| A accepté

Réponse apportée
CUDA Quadro series comparison
For MATLAB, unless you are doing strongly divisible operations such as Monte Carlo sampling, you are better off with one really ...

plus de 10 ans il y a | 0

Réponse apportée
Parallelize a code for Schnakenberg model PDE reaction-diffusion
Try reading up on vectorization: * <http://www.mathworks.com/help/matlab/matlab_prog/vectorization.html> * <http://devblogs....

plus de 10 ans il y a | 0

Réponse apportée
reading saved gpuArray data with a non-gpu computer
Unfortunately, this is not possible. To make data visible to a MATLAB that does not have Parallel Computing Toolbox, you must fi...

plus de 10 ans il y a | 1

| A accepté

Réponse apportée
How to gpuArraying the imported data?
It looks like you are assuming that by putting your mesh data on the GPU it will be available directly to the GPU for graphics d...

plus de 10 ans il y a | 0