Answered
Parameters imported using assignin not visible to parfor
PARFOR requires that you pass all variables explicitly - as you have discovered, it does not understand constructs like ASSIGNIN...

12 years ago | 0

| accepted

Answered
I am new to use parfor, I have searched for the condition but I still can't understand what to do in order to make it work.
To run a PARFOR loop, all variables that you wish to produce as results from the loop must either be _sliced_ or _reductions_. A...

12 years ago | 0

| accepted

Answered
Sliced variables in parfor loop (restricted indexing)
Although you are dividing up 'p' so that each loop iteration works on a different piece, unfortunately you aren't doing it in th...

12 years ago | 2

| accepted

Answered
parfor with wrong indexing. Trying to assign output to parts of vector
At the moment, the PARFOR loop is not order-independent because of the way that you're assigning into "xi". PARFOR loops are req...

12 years ago | 0

| accepted

Answered
can't find GPU in MJS
The MJS workers cannot access the GPU because they're running as a Windows Service, and you need to use the TCC mode driver to a...

12 years ago | 1

| accepted

Answered
Implementing batched mldivide on the GPU
Unfortunately, as you have found PAGEFUN does not yet support batched MLDIVIDE. You might be able to use the <http://www.mathwor...

12 years ago | 1

| accepted

Answered
How to make data persist on workers between calls to parfor?
This looks like it might be a good case for the <http://www.mathworks.com/matlabcentral/fileexchange/31972-worker-object-wrapper...

12 years ago | 0

Answered
Programmatically pre-detect number of cores available to Parallel Computing Toolbox
Try this: c = parcluster('local'); % build the 'local' cluster object nw = c.NumWorkers % get the number of w...

12 years ago | 3

| accepted

Answered
Recursion Limit inside parfor
I would try calling spmd set(0, 'RecursionLimit', N); end as this will ensure it executes on all workers...

12 years ago | 0

| accepted

Answered
Can I use GPU computing inside of a parfor loop?
You do not need 8 separate graphics cards even if you have 8 workers using gpuArrays simultaneously. (In fact, I'm not even sure...

12 years ago | 2

| accepted

Answered
Configire 'local' to run on more than 4 workers
Please note that in R2014a the limit on the number of local workers was removed. See <http://www.mathworks.com/help/distcomp/rel...

12 years ago | 1

Answered
MDCS for n>12 parfor on local multicore computer
Please note that in R2014a the limit on the number of local workers was removed. See <http://www.mathworks.com/help/distcomp/rel...

12 years ago | 1

Answered
how to convince mathworks to increase number of parallel workers?
Please note that in R2014a the limit on the number of local workers was removed. See <http://www.mathworks.com/help/distcomp/rel...

12 years ago | 1

| accepted

Answered
How many MATLAB workers can I open on my client machine?
Please note that in R2014a the limit on the number of local workers was removed. See <http://www.mathworks.com/help/distcomp/rel...

12 years ago | 1

Answered
Optimization using GA and FMINCON with more than 12 processors
Please note that in R2014a the limit on the number of local workers was removed. See <http://www.mathworks.com/help/distcomp/rel...

12 years ago | 1

Answered
How to have more than 12 workers
Please note that in R2014a the limit on the number of local workers was removed. See <http://www.mathworks.com/help/distcomp/rel...

12 years ago | 2

Answered
The current limit of workers in MATLAB2013b for a local cluster is 12 which doesn't reach the server limit of 32 logical cores. A MDCS license for 32 workers costs more than the server hardware. Will MathWorks correct these disparities in the future?
Please note that in R2014a the limit on the number of local workers was removed. See <http://www.mathworks.com/help/distcomp/rel...

12 years ago | 2

| accepted

Answered
Trouble with accessing variable in parfor loop
The problem here is that unfortunately the PARFOR analysis is based on the text of your program, and in this case the analysis c...

12 years ago | 2

| accepted

Answered
Should i install matlab on clusters for parallel computing toolbox?
As well as Parallel Computing Toolbox, you need MATLAB Distributed Computing Server to run on you 'slave' machines. There are in...

12 years ago | 0

Answered
Passing existing COM server to parallel cluster
You need to use actxGetRunningServer on the workers rather than passing the server object directly, like so: >> e = actxser...

12 years ago | 0

Answered
Different behaviour of rng() and rand() in parfor-loop
The client and the workers are set up to use different random generators, so this is expected. This is covered <http://www.mathw...

12 years ago | 0

| accepted

Answered
Unable to start the parallel computing on a linux machine
It looks like you have an 'mpiLibConf.m' on your path that is trying to pick up Intel's MPI. This will not work with the local c...

12 years ago | 0

Answered
PARFOR behavior sensitive to comments???
Unfortunately what I think you're seeing there is a limitation of the way function handles are created. If you use the <http://w...

12 years ago | 1

| accepted

Answered
Fastest Possible Way to convert a table containing Only 2 strings to numbers
You can do this with UNIQUE or ISMEMBER. values = {'Het', '', '', 'Het', '', 'Hom'; '', 'Het', 'Hom', '', '', ''} %...

12 years ago | 0

| accepted

Answered
Split a matrix on parallel computing with spmd
You could do that using <http://www.mathworks.co.uk/help/distcomp/execute-simultaneously-on-multiple-data-sets.html distributed ...

12 years ago | 1

| accepted

Answered
how to configure more than four workers on matlab session?
Note that R2013b allows up to 12 local workers, so you might wish to upgrade.

12 years ago | 0

Answered
can we use multitasking on matlab
You can run up to 12 local workers (assuming R2013b) simultaneously using Parallel Computing Toolbox. You need to <http://www.ma...

12 years ago | 0

| accepted

Answered
Waitbar in a spmd block in a GUI?
In R2013b, there's the new function <http://www.mathworks.co.uk/help/distcomp/parfeval.html PARFEVAL> which lets you program the...

12 years ago | 1

| accepted

Answered
Problem with spmd when data sizes increase
The problem here is that labSend is permitted to block until the corresponding labReceive is posted if the message is "too large...

12 years ago | 2

| accepted

Answered
Configure L1 cache / shared memory size?
You should be able to configure this by using a <http://www.mathworks.co.uk/help/distcomp/run-mex-functions-containing-cuda-code...

12 years ago | 1

| accepted