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 #5809] error loading model: unable to allocate backend buffer when model size is > VRAM with multiple GPUs #65657

New Issue
Closed
opened 2026-05-03 22:06:22 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-05-03 22:06:22 -05:00
Owner
Copy Link

Originally created by @rafal11ck on GitHub (Jul 20, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/5809

What is the issue?

ollama run llava:34b write me a poem

Error: llama runner process has terminated: signal: aborted (core dumped) error loading model: unable to allocate backend buffer
llama_load_model_from_file: exception loading model

Hardware

System has 2 discrete GPUs:

  • AMD RX 7600 XT (16 GB)
  • nvidia 1050 TI (4 GB)

RAM: 48 GB
CPU: AMD 7600X

struggle

I tried to manipulate CUDA_VISIBLE_DEVICES and HIP_VISIBLE_DEVICES envvars. Setting either to -1 makes ollama run with GPU that's left.

Logs:
both.txt
amd_only.txt
nvidia_only.txt

part of both.txt

ggml_backend_cuda_buffer_type_alloc_buffer: allocating 15863.15 MiB on device 0: cudaMalloc failed: out of memory
llama_model_load: error loading model: unable to allocate backend buffer
llama_load_model_from_file: exception loading model
terminate called after throwing an instance of 'std::runtime_error'
  what():  unable to allocate backend buffer
time=2024-07-20T10:45:15.144+02:00 level=ERROR source=sched.go:480 msg="error loading llama server" error="llama runner process has terminated: signal: aborted (core dumped) error loading model: unable to allocate backend buffer"

This happens on every model that I tried and those size > 16GB when both are available.

OS

Linux

GPU

Nvidia, AMD

CPU

AMD

Ollama version

ollama version is 0.2.1

Originally created by @rafal11ck on GitHub (Jul 20, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/5809 ### What is the issue? `ollama run llava:34b write me a poem` ``` Error: llama runner process has terminated: signal: aborted (core dumped) error loading model: unable to allocate backend buffer llama_load_model_from_file: exception loading model ``` # Hardware System has 2 discrete GPUs: - AMD RX 7600 XT (16 GB) - nvidia 1050 TI (4 GB) RAM: 48 GB CPU: AMD 7600X # struggle I tried to manipulate `CUDA_VISIBLE_DEVICES` and `HIP_VISIBLE_DEVICES` envvars. Setting either to `-1` makes ollama run with GPU that's left. Logs: [both.txt](https://github.com/user-attachments/files/16319474/both.txt) [amd_only.txt](https://github.com/user-attachments/files/16319475/amd_only.txt) [nvidia_only.txt](https://github.com/user-attachments/files/16319476/nvidia_only.txt) part of both.txt ``` ggml_backend_cuda_buffer_type_alloc_buffer: allocating 15863.15 MiB on device 0: cudaMalloc failed: out of memory llama_model_load: error loading model: unable to allocate backend buffer llama_load_model_from_file: exception loading model terminate called after throwing an instance of 'std::runtime_error' what(): unable to allocate backend buffer time=2024-07-20T10:45:15.144+02:00 level=ERROR source=sched.go:480 msg="error loading llama server" error="llama runner process has terminated: signal: aborted (core dumped) error loading model: unable to allocate backend buffer" ``` This happens on every model that I tried and those size > 16GB when both are available. ### OS Linux ### GPU Nvidia, AMD ### CPU AMD ### Ollama version ollama version is 0.2.1
GiteaMirror added the memorybug labels 2026-05-03 22:06:22 -05:00
GiteaMirror commented 2026-05-03 22:06:42 -05:00
Author
Owner
Copy Link

@rafal11ck commented on GitHub (Jul 20, 2024):

Updated to 0.2.7 same result.

<!-- gh-comment-id:2241066509 --> @rafal11ck commented on GitHub (Jul 20, 2024): Updated to 0.2.7 same result.
GiteaMirror commented 2026-05-03 22:06:43 -05:00
Author
Owner
Copy Link

@julioarruda commented on GitHub (Jul 25, 2024):

same here

<!-- gh-comment-id:2251287086 --> @julioarruda commented on GitHub (Jul 25, 2024): same here
GiteaMirror commented 2026-05-03 22:06:44 -05:00
Author
Owner
Copy Link

@rafal11ck commented on GitHub (Feb 4, 2025):

Closing as stale, but seems to be kinda resolved

ollama version is 0.5.7

  • it now uses one GPU AMD RX 7600 XT one.
  • if it doesn't fit it offloads to CPU.
<!-- gh-comment-id:2633318397 --> @rafal11ck commented on GitHub (Feb 4, 2025): Closing as stale, but seems to be _kinda_ resolved ollama version is 0.5.7 - it now uses one GPU AMD RX 7600 XT one. - if it doesn't fit it offloads to CPU.
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 bug memory
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#65657
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?