Beantwortet
What's wrong with this parfor loop?
In this case, the fix is simply to remove the increment from the inner |for| loop, like so: N = 10; T = 3; A = zeros(...

mehr als 10 Jahre vor | 1

| akzeptiert

Beantwortet
Reduction function called in a parfor loop cannot have more than two variables?
The problem here is to do with both the number of arguments to your function, and how you're using it with respect to the output...

mehr als 10 Jahre vor | 0

| akzeptiert

Beantwortet
Using getsnapshot in parfor function
This error appears to be misleading - I _think_ the problem is that the object |vid| is not being transferred to the workers cor...

mehr als 10 Jahre vor | 1

| akzeptiert

Beantwortet
Error using rand function with gpu suport
The ability to call |rand(size, type, 'gpuArray')| was added in R2014a. This is mentioned in the <http://www.mathworks.com/help/...

mehr als 10 Jahre vor | 0

Beantwortet
Parfoor loop with 3D matrix
I think the problem here is with the way you're using |signal_filt|. You're using multiple forms of indexing, and the |parfor| r...

mehr als 10 Jahre vor | 0

| akzeptiert

Beantwortet
I want to compile .cu code on my matlab 2015a using 'mexcuda' but i am getting errors as shown
The function |mexcuda| was introduced in R2015b. You'll need to follow the instructions in the R2015a documentation: http://www....

mehr als 10 Jahre vor | 1

| akzeptiert

Beantwortet
Using CUDA 6 for GPU computations (R2015b + parallel toolbox)
Unfortunately, there are at least two different numbers here: 1. Your device has a *CUDA Compute Capability* of 3.5. This ref...

mehr als 10 Jahre vor | 1

Beantwortet
How to efficiently use spmd to load multiple files and perform execution?
Under some circumstances, the "idle" workers can consume CPU resources while waiting for the other workers to get to the end of ...

mehr als 10 Jahre vor | 0

Beantwortet
Parallel computing toolbox and multiple cpu's
The placement of the worker processes is entirely determined by your operating system. So, in practice, PCT local workers will a...

mehr als 10 Jahre vor | 0

Beantwortet
CUDA fft with prime number lengths
Thankyou for reporting this problem. In R2015b, this transform actually fails with an error, so I suspect that in R2015a the tra...

mehr als 10 Jahre vor | 0

Beantwortet
Parpool Fail 2015a HPC
This looks like your machine ran out of resources while trying to start up the workers. Do you have any |ulimit| in effect?

mehr als 10 Jahre vor | 0

Beantwortet
Using arrayfun is giving me an error
|arrayfun| is designed for purely element-wise operations. Your code appears to be attempting to operate on the whole value of |...

mehr als 10 Jahre vor | 0

Beantwortet
Nested functions with GPU does not support FFT2
|arrayfun| with |gpuArray| arguments is primarily designed for *element-wise* computation. Therefore, the functions that you can...

mehr als 10 Jahre vor | 0

Beantwortet
Parfor with COMSOL object
Outputs from a |parfor| loop need to be either _sliced_ outputs, or _reductions_. In this case, a sliced output is probably most...

mehr als 10 Jahre vor | 0

| akzeptiert

Beantwortet
CPU vs. GPU : Can I estimate the improvement by the provided code?
MATLAB will see the K80 device as two separate CUDA devices. Each MATLAB process can access only one device - so to take full ad...

mehr als 10 Jahre vor | 0

| akzeptiert

Beantwortet
Find Peaks In a Faster Way
Unfortunately, |findpeaks| does not currently support |gpuArray| inputs. You might be able to create your own simplified version...

mehr als 10 Jahre vor | 0

| akzeptiert

Beantwortet
Parfor on increasing array
|parfor| does support increasing arrays in this way, even if it is not necessarily desirable. To do that though, you must use |[...

mehr als 10 Jahre vor | 0

Beantwortet
Why complex arrays take twice as much memory on GPU than on CPU ?
Based on the prior comments, I think I understand the problem now. Complex arrays on the GPU take up the same amount of memory a...

mehr als 10 Jahre vor | 2

Beantwortet
Cell arrays and multiple GPUs computing
You could use |spmd| or |parfor| for this, but beware that this will involve making _copies_ of the data onto the GPUs used by t...

mehr als 10 Jahre vor | 0

Beantwortet
How best to parallelize this simple code/algorithm?
For the code as written, since |jj| takes on each value in turn, you can simply flatten this into a single |parfor| loop, unless...

mehr als 10 Jahre vor | 0

Beantwortet
For GPU computing, dedicated memory size will be important or not?
The amount of memory available in your GPU should not affect how quickly it performs calculations (all other things being equal)...

mehr als 10 Jahre vor | 0

| akzeptiert

Beantwortet
Undefined function 'svd' for distributed matrix
Firstly, you should construct your |distributed| array directly on the workers to avoid building the large array at the client, ...

fast 11 Jahre vor | 1

| akzeptiert

Beantwortet
fails to start the parpool
It looks like your system is running out of memory while trying to launch that many workers. How much RAM do you have on your sy...

fast 11 Jahre vor | 0

Beantwortet
Is there any function to test if a mxGPUArray is empty or not?
There isn't a directly equivalent method, but you can call bool const isEmpty = (mxGPUGetNumberOfElements(myArray) == mwSiz...

fast 11 Jahre vor | 1

Beantwortet
How do I set the path for Matlab workers to access packages when using a parallel pool?
It's a little tricky from your description to work out exactly what's gone wrong here, but I think it ought to be possible to re...

fast 11 Jahre vor | 1

Beantwortet
Matlab link to new CUDA toolkit
The version of the CUDA toolkit used by a particular release of MATLAB is fixed, and cannot be changed. Later releases of MATLAB...

fast 11 Jahre vor | 0

Beantwortet
How to read file sequentially in Parfor loop?
To get finer-grained control over the ordering of parallel operations, you can use <http://www.mathworks.com/help/distcomp/spmd....

fast 11 Jahre vor | 0

Beantwortet
GPU or Intel Xenon Phi Coprocessor?
By default, Parallel Computing Toolbox uses as many workers as you have real cores on your system. The upper limit of 512 worker...

fast 11 Jahre vor | 0

Beantwortet
Preserving gpuArray data across Multiple GPUs
Each MATLAB process can access only a single GPU at a time. The only way to preserve data when switching GPU devices is to call ...

fast 11 Jahre vor | 0

Beantwortet
max function in gpu error
On the GPU, |arrayfun| supports only the elementwise version of |min| and |max| - i.e. the two input argument case. In this case...

fast 11 Jahre vor | 1