Answered
Retrieving data from createtask is very slow
The time taken for fetching outputs is driven by the disk access speed, and the speed of MATLAB's own SAVE command. I ran the fo...

12 years ago | 0

Answered
Use whole cpu for my n^2*2^n code
PMODE executes the same line of code on each worker - you need to make that line of behave differently on each worker by using d...

12 years ago | 0

Answered
Random Numbers in Parallel Environment
Parallel Computing Toolbox workers very carefully set their random number generator state to the same state each time a pool is ...

12 years ago | 2

Answered
Controlling new parallel feature "Automatically create a parallel pool when parallel keywords are executed"
Unfortunately, there's no API to modify the parallel preferences. I think you have essentially two options here, and these are: ...

12 years ago | 0

| accepted

Answered
How do I suppress this warning from spmd?
I believe you can only hit this warning if you don't ensure that all workers call labBroadcast together. So, rather than suppres...

12 years ago | 0

| accepted

Answered
Torque Cluster Jobs Submitted, Not Executed
Your message is not expected to be printed in that log file - rather, it should show up in the CommandWindowOutput property of y...

12 years ago | 0

Answered
spmd, parfor, createParellel, Slower than the usual code without going in parallel
There are several reasons why your code might not speed up when going parallel. The most likely is that you're using local worke...

12 years ago | 0

Answered
gpuArray loop indexing question
I think part of your underlying problem must be missing here, since in this case, PHI and phi end up the same. Anyway, you could...

12 years ago | 0

| accepted

Answered
How to use a mex compiled .mexw64 function in a parfor loop
Which cluster type are you using? If you're not using the 'local' cluster, then you might need to send the MEX file to the worke...

12 years ago | 0

Answered
How to use a Com server handle in parfor?
As Walter says, each worker is a separate MATLAB process. Also, each variable is transferred to the workers as if it had been sa...

12 years ago | 0

Answered
Sliced scalar assignment in parfor
Unfortunately, this is a limitation of PARFOR. The only workaround for MATLAB releases prior to R2019b is not to use scalar expa...

12 years ago | 0

| accepted

Answered
How can I use a parfor loop in matlab for a simulink simulation using structs?
There's lots of information about using Simulink inside PARFOR <http://www.mathworks.co.uk/help/simulink/ug/running-parallel-sim...

12 years ago | 0

Answered
nested loops in parfor, indexing
Unfortunately, PARFOR's static analysis cannot tell that you are accessing 'solution' in an order-independent "sliced" manner. Y...

12 years ago | 0

| accepted

Answered
multi-level parallelism in Matlab (2013a)?
As you've discovered, you cannot nest parallelism with PARFOR. Depending on the amount of data transfer etc., you may find that ...

12 years ago | 0

| accepted

Answered
i am trying to get emg and force data simultaneously using parallel computing tool box but i am unable to read the data???
The data contained within a Composite is retained on the workers. Therefore if you close the pool, the data goes away. So, you s...

12 years ago | 1

| accepted

Answered
Error: Transparency violation error.
Transparency violations occur when you try to add variables to the body of the PARFOR loop in a way that isn't clear from the te...

12 years ago | 2

Answered
How do I reduce the memory imprint due to a GPU array?
This is unfortunately likely to be due to loading all the GPU support libraries. These are quite large, and all get loaded when ...

12 years ago | 1

| accepted

Answered
How can I set 32 workers performing parfor?
You need to set the HPC cluster as your default, or you need to name the cluster when opening matlabpool, like so: matlabpo...

12 years ago | 0

Answered
Matlabpool -- can't make it work
Thanks for the detailed reproduction. In this case, MATLAB has already been able to run your code in a multithreaded manner. Thi...

12 years ago | 1

Answered
Backslash operator on GPU
Sparse arrays are now supported on the GPU, here's a page describing the support: https://www.mathworks.com/help/parallel-comput...

12 years ago | 0

| accepted

Answered
Matlab doesn't recognize parpool function
PARPOOL was introduced in R2013b to replace MATLABPOOL. In R2013a and earlier, use <http://www.mathworks.co.uk/help/releases/R20...

12 years ago | 5

Answered
Saving files within parfor loop with every iteration
I presume the problem is that saving data to the disk is causing clashes. You need to make each worker write its results to a se...

12 years ago | 1

Answered
Getting gpuArray output type not currently implemented
The problem here is that 'slice' is not a gpuArray, and so the arrayfun call is executing on the CPU. The problem occurs because...

12 years ago | 0

| accepted

Answered
Matlabpool Maximum Number of Local Workers on one computer - Parallel Computing
The number of local workers available with Parallel Computing Toolbox and no MDCS has changed. When introduced, the limit was 4;...

12 years ago | 2

| accepted

Answered
Behavior of parfeval when its input arguments get updated?
The updates to 'x' will not be seen by the workers. When you call PARFEVAL, the input arguments and function are all packaged up...

12 years ago | 1

| accepted

Answered
Parallel loop variable query
PARFOR currently thinks you're re-using values in percent_inc from one iteration of the PARFOR loop to the next. You can fix thi...

12 years ago | 1

| accepted

Answered
Parfor loop problem with SVM classification
If you're using R2013a or earlier, you need to explicitly open a matlabpool before running the PARFOR loop using "matlabpool('lo...

12 years ago | 0

| accepted

Answered
How overloaded functions are implemented on gpu? I.e. how can I set number of threads and thread blocks when I call. GpuArray?
All operations on gpuArray data take place on the GPU. For built-in things like matrix multiplication, the allocation of blocks ...

12 years ago | 1

| accepted

Answered
parfor - Attempt to serialize data which is too large
The 2GB PARFOR transmission limitation was removed from Parallel Computing Toolbox in R2013a.

12 years ago | 0

| accepted

Answered
Parfor cannot serialize data which is too large
The limit on PARFOR data transfers was removed in R2013a. Please upgrade if you can.

12 years ago | 0

| accepted

Load more