Matlab and Nvidia cuda GPU

2 visualizaciones (últimos 30 días)
Federico Manfrin
Federico Manfrin el 3 de Sept. de 2019
Comentada: Walter Roberson el 9 de Sept. de 2019
Hello, I've the following Mathworks product
  • Matlab, Simulink
  • Curve fitting
  • Simscape electrical
Is it possible to gain any simulation speed using a GPU?
I don't want to buy any other add-on from Mathworks, just installing the gpu provide any benefit?
Another way to ask .. is it possible to use GPU capabilities without specific add-on?
Thanks
  3 comentarios
Walter Roberson
Walter Roberson el 3 de Sept. de 2019
I find no evidence that any of the Simscape products will use the GPU.
Federico Manfrin
Federico Manfrin el 6 de Sept. de 2019
Editada: Federico Manfrin el 6 de Sept. de 2019
Hi Walter, It should be very nice if Simscape blocks uses gpuarray functionalities, and other gpu computing feature, behind the scenes, to improve simulation speed for cuda gpu computer. At least for those who have parallel toolbox .. better for anyone :-)
Hope they will, in the future

Iniciar sesión para comentar.

Respuesta aceptada

Jonas
Jonas el 6 de Sept. de 2019
You can only run calculations on a GPU using the Parallel Computing Toolbox.
  4 comentarios
Jonas
Jonas el 9 de Sept. de 2019
Instead of using a GPU to speed up Simscape simulations, you can use an FPGA to simulate Simscape models. Please see this presentation regarding this topic using an example in the Simscape Electrical toolbox. Depending on the complexity of your model and available hardware, you can run Simscape models in real-time.
Walter Roberson
Walter Roberson el 9 de Sept. de 2019
Communication Toolbox can use gpu, and Deep Learning, and I think I noticed one other.

Iniciar sesión para comentar.

Más respuestas (0)

Categorías

Más información sobre Model Statistics en Help Center y File Exchange.

Etiquetas

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by