OLLAMA Python, Why can't I use all the threads on the CPU? #89

Open
opened 2026-02-15 16:27:56 -05:00 by yindo · 1 comment
Owner

Originally created by @HamzaYslmn on GitHub (May 23, 2024).

image

I analyzed the problem in depth. I get faster responses when I use the terminal, something is wrong with Python. Just use E-cores and its too slow.

I HAVE:

i9 - 13980HX
24 Core - 32 Thread

Originally created by @HamzaYslmn on GitHub (May 23, 2024). ![image](https://github.com/ollama/ollama/assets/78810304/452ca86f-941f-4ff7-b3bf-22c39a3e3c24) I analyzed the problem in depth. I get faster responses when I use the terminal, something is wrong with Python. Just use E-cores and its too slow. I HAVE: i9 - 13980HX 24 Core - 32 Thread
Author
Owner

@HamzaYslmn commented on GitHub (May 24, 2024):

When I test with the GPU, there is a 30% speed difference between running from Python and running from Terminal. Why is this problem happening?

@HamzaYslmn commented on GitHub (May 24, 2024): When I test with the GPU, there is a 30% speed difference between running from Python and running from Terminal. Why is this problem happening?
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: ollama/ollama-python#89