Community Profile

photo

Joss Knight


MathWorks

341 total contributions since 2013

Although I cannot be contacted directly, if you would like to ask me a question all you have to do is mention "GPU" somewhere in your MATLAB Answers question.

Joss Knight's Badges

  • 36 Month Streak
  • Knowledgeable Level 4
  • Pro
  • Revival Level 2
  • First Answer

View details...

Contributions in
View by

Answered
Unexpected speed decrease of 2D Fourier Transform on GPU when iFFTed
I modified your code inserting wait(gpuDevice) before each tic and toc and got a much more sensible graph: The GPU runs async...

16 days ago | 0

| accepted

Answered
How can I accelerate deep learning training using GPU?
Your mini-batch size is far too small. You're not going to get any benefit of GPU over CPU with that little GPU utilisation. Inc...

22 days ago | 0

Answered
GPU utilization is not 100%.
Your question is very hard to answer in it's current form. You want to know why GPU utilisation is not 100%? The answer is, bec...

24 days ago | 0

Answered
Why is arrayfun for GPU slower than normal operations
You are misunderstanding the use of arrayfun for gpuArray. Combine all those operations into a single function. temp00 = arrayf...

27 days ago | 3

| accepted

Answered
GPU parallel computing error with Quadro RTX 5000
I'm amazed this works at all, frankly, since 2012 is about 5 generations of GPU architecture ago. Anyway, there are known issues...

3 months ago | 0

| accepted

Answered
Load of Transfer Learning on GPU
Your problem is this line: imds.ReadFcn = @(loc)imresize(imread(loc),inputSize); You should remove it and instead use an augme...

4 months ago | 0

Answered
Can I use parallel computing to train a DNN?
No, you can't use parallel training for a sequence network, sorry.

4 months ago | 0

| accepted

Answered
Odd error comes during training deep learning model on GPU
Nearly always this error is a kernel timeout. Use Windows regedit and follow the instructions on this page to disable TDR by set...

5 months ago | 0

Answered
Creating a convolutional neural network that outputs an image.
You have a DAGNetwork which means you cannot pass a layer index into activations, you have to pass a name: res = activations(ne...

5 months ago | 1

| accepted

Answered
How to control which GPUs and CPUs get which tasks during multiple calls to trainNetwork?
The computation on the GPU is so much faster than on the CPU for a typical Deep Learning example that there are only disadvantag...

5 months ago | 0

| accepted

Answered
Why system need to wait gpudevice complete?
The fundamental problem is that GPU execution is asynchronous so the point where you think all the time is being spent isn't act...

5 months ago | 0

| accepted

Answered
Support for HIP and/or OpenCL GPU acceleration
MathWorks regularly reviews its policies on GPU support. Where there is a market, a robust, stable, performance portable languag...

5 months ago | 0

Answered
How to generate mex for openCV's 'imdecode' for CPU and GPU?
You can just download and use OpenCV yourself, and link in whatever libraries you need on the compile line. This blog post gives...

5 months ago | 0

Answered
Does MATLAB 2018b Support NVIDIA GeForce 2080 Ti RTX For Creating/Training/Implementing Deep Learning Neural Networks, And, If So, Then Which Version Of CUDA Should I Use?
Yes, you can. Use the following workaround to avoid annoying warnings and errors: warning off parallel:gpu:device:DeviceLibsNee...

5 months ago | 0

Answered
Parallel Computing in Neural Networks is not using all the workers in 2018b?
I believe this is the designed behaviour. If multiple workers were to share the same GPU, you would get a performance reduction,...

6 months ago | 0

Answered
GPU and CPU code: How to do?
This is exactly why MATLAB's GPU support is so awesome! You should only need to insert your 'shim' for gpu data, as OCDER descri...

6 months ago | 2

Answered
Calling CUDA mex functions from functions called via arrayfun
GPU arrayfun supports a restricted set of element-wise operations that it knows how to translate into CUDA device IR. It doesn't...

6 months ago | 0

| accepted

Answered
Declare gpuArray while executing on gpu
GPU arrayfun functions can only do scalar operations. You can declare new scalar variables but you can't create new arrays.

6 months ago | 0

Answered
How can I recompile the GPU libraries ?
This is a known issue with CUDA 9.1 and the new Turing generation of GPUs. You don't need to do anything except ignore the initi...

6 months ago | 3

| accepted

Answered
mexcuda not finding CUDA 10.0 even after I specify it via setenv('MW​_NVCC_PATH​',...)
As it says there in the error message, you need CUDA 9.1, not 10.

6 months ago | 0

Answered
Any GPU implementation of k-nearest neighbor search?
knnsearch is supported on the GPU, so just use it!

6 months ago | 0

Answered
GPU arrayfun is so slow, what is going on?
You haven't called GPU arrayfun here, you've called CPU arrayfun and in the arrayfun function you are doing stuff on the GPU. Th...

6 months ago | 1

Answered
Can I process 'fit' with a GPU?
We plan to accelerate the rate at which we provide improved coverage for gpuArray support in MATLAB Toolboxes, including Stats a...

7 months ago | 1

Answered
Can I process 'fit' with a GPU?
No, there isn't, but other options may be adaptable to your problem. https://uk.mathworks.com/matlabcentral/answers/431697-make...

7 months ago | 1

| accepted

Answered
how can I output different array size than input by GPU arrayfun?
You can't output a variable sized output from GPU arrayfun, which would require atomic operations. You're going to have to compu...

7 months ago | 0

| accepted

Answered
How can I fix the CUDNN errors when I'm running train with RTX 2080?
Regarding issues with memory, the Titan XP has 12GB of memory while the RTX 2080 has only 8GB. You'll need to reduce your MiniBa...

7 months ago | 0

Answered
How can I fix the CUDNN errors when I'm running train with RTX 2080?
This a known issue. Before you start anything else run try nnet.internal.cnngpu.reluForward(1); catch ME end That shoul...

7 months ago | 5

| accepted

Answered
Reduction variables on the GPU II and arrayfun: cannot assign to parent function variable?
No, you can only read from uplevel variables, and then only one element at a time. You cannot write to them. That is not the int...

7 months ago | 0

| accepted

Answered
Using reduction variables on the GPU: arrayfun or other options
I suppose it depends on what f is, is it a scalar operation for each element of x? If so you can move your loop over I inside yo...

7 months ago | 0

| accepted

Answered
MATLAB & Cuda 10
It works, although it will erroneously warn that JIT compilation is required (it isn't). Also, there is a known bug with Deep Le...

7 months ago | 1

Load more