Community Profile

photo

Joss Knight

MathWorks

Last seen: 2 days ago Active since 2013

Statistics

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

View badges

Content Feed

View by

Answered
Compatibility Matlab & GPU coder Compute Capability 8.6 RTX 3070
Regarding the gpuBench results: no, MATLAB is definitely not only using half the cores! What you are seeing is the raw performan...

13 days ago | 0

Answered
How much can a graphics card make a difference in neural networks?
Yes, it could make things run 10x faster maybe.

16 days ago | 0

| accepted

Answered
Can MATLAB ResNet50 be run on Google Colab Pro?
I'm not sure how in-guidelines it is to be asking about running Octave as a substitute for MATLAB on this forum! I'm pretty sure...

16 days ago | 1

Answered
BiCGStab is slower on R2021a using GPUs
There is a MATLAB bug report on this known issue: https://www.mathworks.com/support/bugreports/details/2438221 You should make...

1 month ago | 0

Answered
Optimization of loops with deep learning functions
dlconv is designed to work in batch with multiple input channels, multiple filters, and multiple input observations in a single ...

1 month ago | 0

| accepted

Answered
Is R2021a configured to automatically use Apple M1 chip for GPU and Neural Engine?
No. MATLAB only supports NVIDIA GPUs for computation.

2 months ago | 1

| accepted

Answered
Why is pagemtimes slower on GPU than a CPU?
This might be simply because you are running double-precision math on a device designed for single precision operations. gpuBenc...

2 months ago | 0

Answered
Problems in defining a custom Neural network layer
Your custom layer doesn't work for input sequences, and it needs to. So the particular error that is happening (which you can se...

3 months ago | 0

Answered
Cellarray in a dlarray
dlCell = cellfun(@dlarray, myCell, 'UniformOutput', false);

3 months ago | 0

Answered
MATLAB : find a way with GPU NVIDIA to inverse quickly a large matrix of symbolic variables
Sorry, but Symbolic Math in MATLAB does not have GPU support.

4 months ago | 1

Answered
Which version of MATLAB will provide functional (i.e. working) GUI interactions?
You made a pre-emptive request that any response not be evasive and I wanted to make an attempt to answer your original question...

4 months ago | 1

Answered
NVIDIA QUADRO CUDA® Toolkit Version
MATLAB R2021a supports all NVIDIA GPUs as of this date, so feel free to get any card you want! The fastest Quadro I know of righ...

4 months ago | 0

| accepted

Answered
Parallel Computing Toolbox (parfor slower than for, GPU slower than CPU)
1) CPU/parfor: How can I measure the time transfer when doing parfor (since parfor is slower than for when calling to a part of ...

4 months ago | 0

Answered
GPU Coder vs. ONNXRuntime, is there a difference in inference speed?
You can compile your network for TensorRT using GPU Coder if that's your intended target, no need to go through ONNX. I don't b...

4 months ago | 0

Answered
Why is my code running slower on the GPU?
Increasing the batch size alone cannot improve convergence in a simple MLP, you need to match it with an increase to the learnin...

4 months ago | 0

| accepted

Answered
Can I tell Matlab not to use contiguous memory?
You might want to ask yourself why you need so many variables in your workspace at once and whether you couldn't make better use...

4 months ago | 2

Answered
How to plot heatmap using gpuArray?
Sorry, scatteredInterpolant doesn't support gpuArray.

4 months ago | 0

Answered
Can I hold 2 batches of dlnetwork gradients and update network parameters in 1 operation?
Yes, absolutely, just sum the gradients until your batch size is the size you want, then update the model. The principle is exac...

5 months ago | 0

| accepted

Answered
Unexpected error calling cuDNN: CUDNN_STATUS_BAD_PARAM.
After some investigation (see thread below), this problem seems to be limited to RTX 3080 and 3070 and Linux. It can be worked a...

5 months ago | 4

| accepted

Answered
Image being labelled something else than in database
You have called classify using the variable net. But the network you trained is called netTransfer. %Train the network that con...

5 months ago | 1

| accepted

Answered
How to print in cuda kernel when writing mex files
For convenience, printf is redirected by mex.h so that output will appear in the command window. You can use printf, but you sho...

5 months ago | 0

| accepted

Answered
For indexing in gpuArray, how can I improve my code?
Firstly, use pdist2. Secondly, say no to loops: Y = reshape(obj.x_train',1,[],num_train); delta = X-Y; dists = squeeze(sqrt(...

5 months ago | 0

| accepted

Answered
Why doesn't my CNN fit into the memory of the GPU?
VGG16 is a 1GB model, if you inflate it to 3-D you're going to have very serious memory pressure. More to the point, the error y...

5 months ago | 0

| accepted

Answered
large training dataset leads to error "out of memory. Type "help memory" for your options. caused by: out of memory. Type "help memory " for your options".
Have you followed the following advice from the example?: "The images in the CamVid data set are 720 by 960 in size. Image size...

5 months ago | 0

Answered
How can we train using gpu instead of cpu ?
By default your GPU will be used if you have one. To force it, set the ExecutionEnvironment training option to 'gpu'.

5 months ago | 0

| accepted

Answered
mexcuda in ubuntu error
Try mexcuda -lcublas GEMM_GPU.cu

5 months ago | 0

| accepted

Answered
mexcuda in ubuntu error
Your declaration const mwSize dims[] = {numCRows,numCCols,batch_count}; Is casting away the signed-ness of your variable numCR...

5 months ago | 0

Answered
Inconsistent results when using a Cuda Kernel in Matlab
You are overwriting the contents of wReal and wImag inside your kernel, but every thread with the same x grid position is readin...

5 months ago | 0

| accepted

Answered
Could I use the `mxGPUArray` format data in the cuda kernel function?
No. You have to get the GPU data pointer using the mxGPUGetData function.

6 months ago | 0

Answered
How `gpuArray` save sparse matrix when running Preconditioned conjugated gradient?
gpuArray currently stores sparse matrices internally in CSR format. This matches the NVIDIA cusparse routines that are used for ...

6 months ago | 1

| accepted

Load more