0
$\begingroup$

I have two processors (two different physical processors) with base frequencies 3.2 Ghz and 4.0 Ghz.
I want to ask if there exist a method which I can use to the 3.2 Ghz processor for ML acceleration ?
I don't have a good gpu for this task.
Will it be slower than a GPU ?

Improve this question
$\endgroup$
1
  • $\begingroup$ It's not "acceleration", it is just running ML on a CPU. "acceleration" means something faster than a CPU. But surely if you do have two CPUs you can use both. $\endgroup$
    – Criticizing Israel not allowed
    Commented Jan 5, 2023 at 0:40

1 Answer 1

Reset to default
1
$\begingroup$

I hope this fits as an answer to your question

It depends on what type of ML you are trying to do.

For example, I don't really know how to parralelize reinforcement learning techniques (most likely there is a way to parralelize at least some of them), but it is simple to parallelize a supervized learning technique that uses batching - simply distribute the batch between the two processors, and then aggregate the results.

I can't tell you if this will be faster of slower than using GPU, since it really depends on the GPU as well. But taking into consideration that GPUs are good at those kinds of computations, my guess would be that a GPU could still be faster if your supervised model is large.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Thank you for answer. $\endgroup$
    – Akash
    Commented Sep 23, 2021 at 14:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.