github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 00:22:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #5747] Support to Intel NPU by Intel NPU Acceleration Library #29340

New Issue
Open
opened 2026-04-22 08:06:35 -05:00 by GiteaMirror · 30 comments
GiteaMirror commented 2026-04-22 08:06:35 -05:00
Owner
Copy Link

Originally created by @lordpba on GitHub (Jul 17, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/5747

Originally assigned to: @dhiltgen on GitHub.

Should be great to add support to the new and future Intel Neural Processing Units

There is already a library for this https://github.com/intel/intel-npu-acceleration-library and i.e. for Phi-3 is great.

I am sure that NPUs will be everywhere, and they will be a viable alternative to CUDA and Nvidia GPUs, imho.

Originally created by @lordpba on GitHub (Jul 17, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/5747 Originally assigned to: @dhiltgen on GitHub. Should be great to add support to the new and future Intel Neural Processing Units There is already a library for this https://github.com/intel/intel-npu-acceleration-library and i.e. for Phi-3 is great. I am sure that NPUs will be everywhere, and they will be a viable alternative to CUDA and Nvidia GPUs, imho.
GiteaMirror added the intelfeature request labels 2026-04-22 08:06:35 -05:00
GiteaMirror commented 2026-04-22 08:06:36 -05:00
Author
Owner
Copy Link

@liyimeng commented on GitHub (Oct 28, 2024):

Any progress on this? ;)

<!-- gh-comment-id:2440463964 --> @liyimeng commented on GitHub (Oct 28, 2024): Any progress on this? ;)
GiteaMirror commented 2026-04-22 08:06:36 -05:00
Author
Owner
Copy Link

@Pospitch commented on GitHub (Nov 28, 2024):

I just got new laptop - ASUS Zenbook S 14 with Intel Core Ultra 7 258V and I was expecting Ollama will use it's NPU by default, so I was surprised it's not using NPU at all. I would also like to know if there is any progress on this, because using CPU only when there is 47 TOPS NPU is just sad.

<!-- gh-comment-id:2506558926 --> @Pospitch commented on GitHub (Nov 28, 2024): I just got new laptop - ASUS Zenbook S 14 with Intel Core Ultra 7 258V and I was expecting Ollama will use it's NPU by default, so I was surprised it's not using NPU at all. I would also like to know if there is any progress on this, because using CPU only when there is 47 TOPS NPU is just sad.
GiteaMirror commented 2026-04-22 08:06:37 -05:00
Author
Owner
Copy Link

@gitsupratim commented on GitHub (Jan 7, 2025):

Even alienwares are packing this NPU these days. Good thing if you can use NPUs on standard LLMs frameworks. Please see if you can enable the support for Intel NPU!

<!-- gh-comment-id:2575346733 --> @gitsupratim commented on GitHub (Jan 7, 2025): Even alienwares are packing this NPU these days. Good thing if you can use NPUs on standard LLMs frameworks. Please see if you can enable the support for Intel NPU!
GiteaMirror commented 2026-04-22 08:06:37 -05:00
Author
Owner
Copy Link

@JonasPerssonSE commented on GitHub (Jan 28, 2025):

Any progress?

<!-- gh-comment-id:2618223330 --> @JonasPerssonSE commented on GitHub (Jan 28, 2025): Any progress?
GiteaMirror commented 2026-04-22 08:06:38 -05:00
Author
Owner
Copy Link

@keruru-amuri commented on GitHub (Jan 29, 2025):

Would be nice to have this feature

<!-- gh-comment-id:2621626858 --> @keruru-amuri commented on GitHub (Jan 29, 2025): Would be nice to have this feature
GiteaMirror commented 2026-04-22 08:06:38 -05:00
Author
Owner
Copy Link

@sammeowww commented on GitHub (Feb 1, 2025):

+1 wish to have this feature as well ;)

<!-- gh-comment-id:2628984065 --> @sammeowww commented on GitHub (Feb 1, 2025): +1 wish to have this feature as well ;)
GiteaMirror commented 2026-04-22 08:06:39 -05:00
Author
Owner
Copy Link

@Morigun commented on GitHub (Feb 9, 2025):

+1

<!-- gh-comment-id:2646136145 --> @Morigun commented on GitHub (Feb 9, 2025): +1
GiteaMirror commented 2026-04-22 08:06:39 -05:00
Author
Owner
Copy Link

@botany233 commented on GitHub (Feb 9, 2025):

+1

<!-- gh-comment-id:2646165792 --> @botany233 commented on GitHub (Feb 9, 2025): +1
GiteaMirror commented 2026-04-22 08:06:39 -05:00
Author
Owner
Copy Link

@WeihanLi commented on GitHub (Feb 9, 2025):

+1

<!-- gh-comment-id:2646166872 --> @WeihanLi commented on GitHub (Feb 9, 2025): +1
GiteaMirror commented 2026-04-22 08:06:40 -05:00
Author
Owner
Copy Link

@zhuxixi commented on GitHub (Feb 10, 2025):

+1

<!-- gh-comment-id:2646869878 --> @zhuxixi commented on GitHub (Feb 10, 2025): +1
GiteaMirror commented 2026-04-22 08:06:40 -05:00
Author
Owner
Copy Link

@i24E commented on GitHub (Feb 14, 2025):

+1

<!-- gh-comment-id:2658185527 --> @i24E commented on GitHub (Feb 14, 2025): +1
GiteaMirror commented 2026-04-22 08:06:41 -05:00
Author
Owner
Copy Link

@UsmanQureshi commented on GitHub (Feb 19, 2025):

+1

<!-- gh-comment-id:2668874798 --> @UsmanQureshi commented on GitHub (Feb 19, 2025): +1
GiteaMirror commented 2026-04-22 08:06:42 -05:00
Author
Owner
Copy Link

@SamuelPalubaCZ commented on GitHub (Feb 19, 2025):

+1

<!-- gh-comment-id:2669510347 --> @SamuelPalubaCZ commented on GitHub (Feb 19, 2025): +1
GiteaMirror commented 2026-04-22 08:06:43 -05:00
Author
Owner
Copy Link

@njduck commented on GitHub (Feb 25, 2025):

+1

<!-- gh-comment-id:2680983042 --> @njduck commented on GitHub (Feb 25, 2025): +1
GiteaMirror commented 2026-04-22 08:06:43 -05:00
Author
Owner
Copy Link

@CRASH-Tech commented on GitHub (Mar 3, 2025):

+1

<!-- gh-comment-id:2693698377 --> @CRASH-Tech commented on GitHub (Mar 3, 2025): +1
GiteaMirror commented 2026-04-22 08:06:44 -05:00
Author
Owner
Copy Link

@crashless commented on GitHub (Mar 5, 2025):

+1

<!-- gh-comment-id:2699676122 --> @crashless commented on GitHub (Mar 5, 2025): +1
GiteaMirror commented 2026-04-22 08:06:45 -05:00
Author
Owner
Copy Link

@JavierChaparroMX commented on GitHub (Mar 23, 2025):

would greatly benefit by this feature being available! +1

<!-- gh-comment-id:2745955570 --> @JavierChaparroMX commented on GitHub (Mar 23, 2025): would greatly benefit by this feature being available! +1
GiteaMirror commented 2026-04-22 08:06:47 -05:00
Author
Owner
Copy Link

@hellerqr commented on GitHub (Mar 29, 2025):

It would be great for laptops and smaller models to use on the go!

<!-- gh-comment-id:2763379635 --> @hellerqr commented on GitHub (Mar 29, 2025): It would be great for laptops and smaller models to use on the go!
GiteaMirror commented 2026-04-22 08:06:48 -05:00
Author
Owner
Copy Link

@DocMAX commented on GitHub (Mar 29, 2025):

There is a ipex-llm ollama version. Isn't that for NPU?

<!-- gh-comment-id:2764272763 --> @DocMAX commented on GitHub (Mar 29, 2025): There is a ipex-llm ollama version. Isn't that for NPU?
GiteaMirror commented 2026-04-22 08:06:49 -05:00
Author
Owner
Copy Link

@cft5059 commented on GitHub (Apr 27, 2025):

+1

<!-- gh-comment-id:2833448803 --> @cft5059 commented on GitHub (Apr 27, 2025): +1
GiteaMirror commented 2026-04-22 08:06:50 -05:00
Author
Owner
Copy Link

@drindt commented on GitHub (May 17, 2025):

This is a feature that I (and probably many others) would love to see implemented as well. Would there be interest in pooling together some funds to sponsor or support the developer for this feature? If enough of us contribute, it might help prioritize its development.

<!-- gh-comment-id:2888169515 --> @drindt commented on GitHub (May 17, 2025): This is a feature that I (and probably many others) would love to see implemented as well. Would there be interest in pooling together some funds to sponsor or support the developer for this feature? If enough of us contribute, it might help prioritize its development.
GiteaMirror commented 2026-04-22 08:06:50 -05:00
Author
Owner
Copy Link

@levi2m commented on GitHub (May 26, 2025):

its so sad to not use my Intel Core Ultra 7 165U npu on ollama... 👍

<!-- gh-comment-id:2908207847 --> @levi2m commented on GitHub (May 26, 2025): its so sad to not use my Intel Core Ultra 7 165U npu on ollama... 👍
GiteaMirror commented 2026-04-22 08:06:51 -05:00
Author
Owner
Copy Link

@thesolomon-tech commented on GitHub (May 26, 2025):

Since the Intel NPU acceleration library is now no longer being maintained, I think it would be best to close this issue in favour of https://github.com/ollama/ollama/issues/2169.

<!-- gh-comment-id:2908437987 --> @thesolomon-tech commented on GitHub (May 26, 2025): Since the Intel NPU acceleration library is now no longer being maintained, I think it would be best to close this issue in favour of https://github.com/ollama/ollama/issues/2169.
GiteaMirror commented 2026-04-22 08:06:52 -05:00
Author
Owner
Copy Link

@zopnote commented on GitHub (Jun 28, 2025):

Also want this feature

<!-- gh-comment-id:3015276888 --> @zopnote commented on GitHub (Jun 28, 2025): Also want this feature
GiteaMirror commented 2026-04-22 08:06:53 -05:00
Author
Owner
Copy Link

@JoseMariaZ commented on GitHub (Jul 4, 2025):

For all those who want to run LLM on Intel NPU, check this link:

https://github.com/JoseMariaZ/Intelexia

<!-- gh-comment-id:3036176031 --> @JoseMariaZ commented on GitHub (Jul 4, 2025): For all those who want to run LLM on Intel NPU, check this link: https://github.com/JoseMariaZ/Intelexia
GiteaMirror commented 2026-04-22 08:06:54 -05:00
Author
Owner
Copy Link

@SiewwenL commented on GitHub (Jul 17, 2025):

Recently, Ollama has integrate with OpenVINO that can leverage Intel hardware accelerators, which include CPU/GPU/NPU. You can find more details here:
https://github.com/openvinotoolkit/openvino_contrib/tree/master/modules/ollama_openvino

<!-- gh-comment-id:3084553805 --> @SiewwenL commented on GitHub (Jul 17, 2025): Recently, Ollama has integrate with OpenVINO that can leverage Intel hardware accelerators, which include CPU/GPU/NPU. You can find more details here: https://github.com/openvinotoolkit/openvino_contrib/tree/master/modules/ollama_openvino
GiteaMirror commented 2026-04-22 08:06:54 -05:00
Author
Owner
Copy Link

@rklec commented on GitHub (Jan 27, 2026):

Also found https://github.com/carloderossi/OllamaWin64NPU-GPU/tree/main that is likely 100% AI generated and claims to just set some parameters for Ollama to make it use the NPU. I tested it and it does not work...

As for https://github.com/carloderossi/OllamaWin64NPU-GPU/tree/main downloading some exe from random Google Drives or so does not really inspire confidence.
Having it integrated in OLLAMA would still be best, I guess.

<!-- gh-comment-id:3804184222 --> @rklec commented on GitHub (Jan 27, 2026): Also found https://github.com/carloderossi/OllamaWin64NPU-GPU/tree/main that is likely 100% AI generated and claims to just set some parameters for Ollama to make it use the NPU. I tested it and it does not work... As for https://github.com/carloderossi/OllamaWin64NPU-GPU/tree/main downloading some exe from random Google Drives or so does not really inspire confidence. Having it integrated in OLLAMA would still be best, I guess.
GiteaMirror commented 2026-04-22 08:06:55 -05:00
Author
Owner
Copy Link

@jclab-joseph commented on GitHub (Feb 19, 2026):

+1

<!-- gh-comment-id:3927284933 --> @jclab-joseph commented on GitHub (Feb 19, 2026): +1
GiteaMirror commented 2026-04-22 08:06:55 -05:00
Author
Owner
Copy Link

@AwaleSagar commented on GitHub (Mar 1, 2026):

+1

<!-- gh-comment-id:3978936684 --> @AwaleSagar commented on GitHub (Mar 1, 2026): +1
GiteaMirror commented 2026-04-22 08:06:56 -05:00
Author
Owner
Copy Link

@valdeirsapara commented on GitHub (Mar 24, 2026):

+1

<!-- gh-comment-id:4122080160 --> @valdeirsapara commented on GitHub (Mar 24, 2026): +1
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label feature request intel
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#29340
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?