Using GPU cores for parallel computing in Tensorflow

Question

I am working on tensorflow-gpu and pyqt5 for an object detection system.

I have developed a project based on neural network model which is trained by tensorflow and it is good but I need to speed up the detection rate.

My GPU is GTX 1060 and CPU is Corei7.

How can I use CUDA cores to divide my computations ?

I have searched a lot of articles and I have asked it several times in *stack** overflow* but there is no response.

How can I use tensorflow-gpu with programming CUDA cores by libraries like PYCUDA or numba or CUPY?

I have asked this question in several manners but I am looking for a right approach to use CUDA for programming GPU cores( GTX1060 has 1280 CUDA cores but Corei7 has 8 core and by giving computations to GPU , program will speed up in a large scale)

pierresegonne · Accepted Answer · 2020-04-06 09:38:19Z

1

See here for installation of tensorflow-gpu. In a script, tensorflow automatically uses GPU if available, but you can check this for more information to check the number of available cores or select some manually etc

answered Apr 6, 2020 at 9:38

pierresegonne

4461 gold badge4 silver badges16 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Amin Darestani Over a year ago

thanks for your reply but the link you have inserted is about how tensorflow can work with more than one GPU , not about how each GPU cores can work with tensorflow.

Collectives™ on Stack Overflow

Using GPU cores for parallel computing in Tensorflow

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related