You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have read the documentation and re-compile llama.cpp using cmake -B build -DGGML_CUDA=ON, however it seems that it can only detect one of my GPU cards. Anyone know why?
$ sunny@hpcnode2:~$ ./llama.cpp/build/bin/llama-cli Llama-3.3-70B-Instruct-F16.gguf -ngl 81 --list-devices
ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 CUDA devices:
Device 0: NVIDIA A800-SXM4-80GB MIG 4g.40gb, compute capability 8.0, VMM: yes
Available devices:
CUDA0: NVIDIA A800-SXM4-80GB MIG 4g.40gb (40448 MiB, 40196 MiB free)
$ sunny@hpcnode2:~$ nvidia-smi
Tue Feb 4 00:57:51 2025
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.161.07 Driver Version: 535.161.07 CUDA Version: 12.2 |
|-----------------------------------------+----------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+======================+======================|
| 0 NVIDIA A800-SXM4-80GB On | 00000000:B7:00.0 Off | On |
| N/A 30C P0 53W / 400W | 37MiB / 81920MiB | N/A Default |
| | | Enabled |
+-----------------------------------------+----------------------+----------------------+
| 1 NVIDIA A800-SXM4-80GB On | 00000000:BD:00.0 Off | On |
| N/A 33C P0 55W / 400W | 1MiB / 81920MiB | N/A Default |
| | | Enabled |
+-----------------------------------------+----------------------+----------------------+
+---------------------------------------------------------------------------------------+
| MIG devices: |
+------------------+--------------------------------+-----------+-----------------------+
| GPU GI CI MIG | Memory-Usage | Vol| Shared |
| ID ID Dev | BAR1-Usage | SM Unc| CE ENC DEC OFA JPG |
| | | ECC| |
|==================+================================+===========+=======================|
| 0 2 0 0 | 21MiB / 40448MiB | 56 0 | 4 0 2 0 0 |
| | 0MiB / 65535MiB | | |
+------------------+--------------------------------+-----------+-----------------------+
| 1 0 0 0 | 1MiB / 81228MiB | 98 0 | 7 0 5 1 1 |
| | 1MiB / 131072MiB | | |
+------------------+--------------------------------+-----------+-----------------------+
+---------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=======================================================================================|
| No running processes found |
+---------------------------------------------------------------------------------------+
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
I have read the documentation and re-compile llama.cpp using
cmake -B build -DGGML_CUDA=ON
, however it seems that it can only detect one of my GPU cards. Anyone know why?Beta Was this translation helpful? Give feedback.
All reactions